A Parallelized PIC/MCC Model with realistic SEY-Based Surface Emission Modeling for Capacitively Coupled Plasmas

This study develops a CUDA C++-based 1D3V PIC/MCC parallel simulation model for capacitively coupled plasmas. The model introduces a novel particle management method that enables efficient addition and deletion of particles in GPU memory, ensuring that all solution processes are executed in parallel on the GPU, thereby avoiding time losses associated with frequent CPU-GPU communication. Furthermore, the proposed model incorporates a realistic secondary electron emission(SEE) model based on secondary electron yield (SEY), rendering the simulation results more physically realistic. The proposed model has been validated against benchmark cases, significantly reducing solution time while ensuring computational accuracy, with the majority of cases completing within several minutes to hours.compared to the CPU, it achieves over 30 times speedup for hundred thousand particles; The pseudo-collision algorithm yields slightly lower electron density than the original Monte Carlo method but exhibits reduced efficiency advantages in parallel architectures; A Unique optimal grid density minimizes runtime by balancing charge assignment and Poisson solving. This model provides an efficient solution for large-scale plasma simulations.