Mechanical Energy Storage for Self-Destructing Motes and Jumping Microrobots

Joseph Greenspun

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2018-112
August 10, 2018

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2018/EECS-2018-112.pdf

Mechanical energy storage has been studied to enable a self-destructing mote and a jumping microrobot. Just as chemical energy stored in batteries can be used for a wide array of devices, so too can mechanical energy stored in beams. In this work various energy storage elements are designed, fabricated, and tested to create these MEMS systems. As privacy and data security become increasingly important, new means of keeping that data safe must be developed. A MEMS system has been created to allow a wirelessly enabled mote to destroy itself on command. To achieve this, a cavity is microfabricated, lled with a silicon etchant, and capped with a fracturable membrane. An energy storage device capable applying 100's of milinewtons of force across distances of 10's of microns was designed and fabricated to fracture these membranes. Additionally, an electrostatic latch was developed to electrically trigger the release of the stored energy. The voltage required to keep this latch closed was reduced using a series of lever arms to amplify the electrostatic force. Two versions of a silicon jumping microrobots were developed as well. The rst microrobot had no active force-producing components and used identical energy storage elements as the self-destructing mote project. In the second microrobot design, the energy storage elements were redesigned and optimized to work with an electrostatic inchworm motor. This motor was combined with a rack and pinion system to create a motor system capable of amplifying the force from a standard inchworm motor by a factor of 10. This microrobot was capable of storing 1.0 nullJ of mechanical energy and jumping 1 mm when its motors were actuated electrically through tethered inputs. When the energy storage elements were loaded manually and latched using one of the inchworm motors, 4.0 nullJ of energy were stored and the microrobot jumped 6.5 mm. Finally, a design and simulation library was created throughout this work speci cally for microrobots. This library, written in MATLAB, can be used to programmatically generate layout les as well as simulation les. While this functionality exists in other software packages, the MATLAB environment enables calculations to be done in-line with the layout. Users can easily add new functions and build upon the existing software. The simulation environment uses a solid body physics simulator to test functionality of microrobots in software before they are fabricated. This helps ensure that new designs work as intended before going through the time intensive and costly process of fabrication in the cleanroom.

Advisor: Kristofer Pister


BibTeX citation:

@phdthesis{Greenspun:EECS-2018-112,
    Author = {Greenspun, Joseph},
    Title = {Mechanical Energy Storage for Self-Destructing Motes and Jumping Microrobots},
    School = {EECS Department, University of California, Berkeley},
    Year = {2018},
    Month = {Aug},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2018/EECS-2018-112.html},
    Number = {UCB/EECS-2018-112},
    Abstract = {Mechanical energy storage has been studied to enable a self-destructing mote and a
jumping microrobot. Just as chemical energy stored in batteries can be used for a wide
array of devices, so too can mechanical energy stored in beams. In this work various energy
storage elements are designed, fabricated, and tested to create these MEMS systems.
As privacy and data security become increasingly important, new means of keeping that
data safe must be developed. A MEMS system has been created to allow a wirelessly enabled
mote to destroy itself on command. To achieve this, a cavity is microfabricated, lled with a
silicon etchant, and capped with a fracturable membrane. An energy storage device capable
applying 100's of milinewtons of force across distances of 10's of microns was designed and
fabricated to fracture these membranes. Additionally, an electrostatic latch was developed
to electrically trigger the release of the stored energy. The voltage required to keep this latch
closed was reduced using a series of lever arms to amplify the electrostatic force.
Two versions of a silicon jumping microrobots were developed as well. The rst microrobot
had no active force-producing components and used identical energy storage elements
as the self-destructing mote project. In the second microrobot design, the energy storage
elements were redesigned and optimized to work with an electrostatic inchworm motor. This
motor was combined with a rack and pinion system to create a motor system capable of
amplifying the force from a standard inchworm motor by a factor of 10. This microrobot
was capable of storing 1.0 J of mechanical energy and jumping 1 mm when its motors
were actuated electrically through tethered inputs. When the energy storage elements were
loaded manually and latched using one of the inchworm motors, 4.0 J of energy were stored
and the microrobot jumped 6.5 mm.
Finally, a design and simulation library was created throughout this work specically for
microrobots. This library, written in MATLAB, can be used to programmatically generate
layout les as well as simulation les. While this functionality exists in other software
packages, the MATLAB environment enables calculations to be done in-line with the layout.
Users can easily add new functions and build upon the existing software. The simulation
environment uses a solid body physics simulator to test functionality of microrobots in software before they are fabricated. This helps ensure that new designs work as intended
before going through the time intensive and costly process of fabrication in the cleanroom.}
}

EndNote citation:

%0 Thesis
%A Greenspun, Joseph
%T Mechanical Energy Storage for Self-Destructing Motes and Jumping Microrobots
%I EECS Department, University of California, Berkeley
%D 2018
%8 August 10
%@ UCB/EECS-2018-112
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2018/EECS-2018-112.html
%F Greenspun:EECS-2018-112