Speaker: Peicheng Yu
Affiliation: UCLA - Ph.D. Candidate
Abstract: In this study, we systematically explore the use of a speedup technique, called Lorentz boosted frame simulation technique, for the Particle-in-cell (PIC) algorithm, with a focus on its application in laser wakefield acceleration (LWFA) simulation. To realize the LWFA simulation in the Lorentz boosted frame, we first developed a theory that explained the source of a robust numerical instability observed in LWFA boosted frame simulations, called numerical Cerenkov instability (NCI), and discussed the methods to eliminate it. These elimination methods include the modification of the Maxwell solvers, plus the corresponding corrections and filtering of the current in the algorithm for the eliminations of specific NCI modes. We then discussed a method to combine the Lorentz boosted frame simulation technique with another speed up technique, called quasi-3D algorithm, to gain unprecedented speed up for the LWFA simulations. The key is the use of a hybrid Maxwell solver in which the plasma drifting direction is solved with higher-order accuracy plus intentional modification of EM dispersion in that direction, while keeping the way Maxwell equation is conventionally solved in other direction(s). We implemented these algorithms in production codes UPIC-EMMA and OSIRIS, and observed unprecedented speedups when simulating various LWFA problems. Besides the LWFA Lorentz boosted frame simulation, we found that NCI elimination schemes can also be used for the simulation of relativistic shock, in which two counter-propagating plasmas are colliding with each other. We showed how applying the NCI elimination scheme leads to very clean shock simulation results, which opens the possibility of tremendous physics advance in this field.
Biography: Peicheng Yu received his BS and MS degrees in Engineering Physics from Tsinghua University, China. In 2010, he joined Professor Warren Mori’s Plasma Simulation Group in Electrical Engineering Department at UCLA to pursue his Ph.D. degree. His research focuses on the Particle-in-cell simulation algorithm for the relativistic plasma drift, Lorentz boosted frame simulation technique, and physics of plasma-based accelerator. During his research at UCLA, Peicheng has authored or co-authored 18 peer reviewed research articles in major journals and conferences.
Date(s) - Apr 20, 2016
1:00 pm - 3:00 pm
E-IV Maxwell Room #57-124
420 Westwood Plaza - 5th Flr. , Los Angeles CA 90095