Paul Christopher Hansen, Research Associate, Stanford University, email@example.com
Lars Thorben Neustock, PhD Candidate, Stanford University, firstname.lastname@example.org
Zach Russell, Collaborator, Ion-Innovations, email@example.com
In this research, we develop algorithms to optimize the shape of physical devices. In photonics and other areas of applied physics, the geometrical shape of a device plays a crucial role in its performance. After calculating the sensitivity of the device performance to the parameters defining its shape (the “design sensitivity”), we efficiently optimize devices in nanophotonics and charged-particle optics such as waveguides, apertures, and electron lensing elements. To calculate the sensitivity, we use the adjoint variable method. First, we model the physical behavior with numerical methods such as FDTD and FEM. Next we differentiate the simulation with respect to design parameters and derive a so-called adjoint system. By solving both the original and adjoint systems we calculate the design sensitivity, without requiring additional system solves for each design parameter as standard finite-difference algorithms do. Additionally, we are investigating algorithms for next-generation shape optimization.
The applications of devices we optimized numerically range from optical tweezers and particle transport, near-field scanning microscopy and optical data storage to improved lithography and X-ray systems.
There is currently no content classified with this term.