*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Julie Kraus
(Advisor: Prof. Rimoli)
will propose a doctoral thesis entitled,
Models for the Non-Linear Response of Tensegrity Meta-materials
On
Thursday, November 17 at 1:30 p.m.
Montgomery Knight 212
Abstract
Tensegrity structures are composed of bars and tethers that have a state where they can be stability stressed without outside constraints. This makes them pliable, controllable, and capable of large deformations. In order to use tensegrities in actual applications it is necessary to have the capability to model them. Current computational modeling methods involve finite element modeling (FEM). FEM is a very powerful tool that can analyze most systems given enough time and computational power. However, for certain applications related to tensegrity structures, the required computational power can become a limiting factor.
In this work we develop computational models to further improve our understanding of tensegrity structures. For the case of pin-jointed tensegrity structures, we use the Elastica theory to derive an analytical model of deformation in imperfect beams. Our model outperforms existing ones and is more accurate over a large range of deformations. We also then utilize finite element simulations to study the behavior of 3D-printed tensegrity metamaterials and demonstrate that the same type of response observed in pin-jointed structures is also held in the frame cases, and that tensegrities have strong potential as structures that can handle high strain prior to failure due to their ability to delocalize failure. Finally, by observing that these models are accurate but computationally expensive, we develop a machine learning based reduced order model to predict the response of tensegrity structures. Our algorithm is based on a frame-indifferent neural network architecture, allowing for the ability to predict the effects of large-scale deformation on tensegrity structures.
Committee
