高温物理气相合成Ag纳米颗粒的Monte Carlo-GPU并行模拟

Parallel Simulation of Gas Synthesis of Ag Nanoparticles at High Temperature With Monte Carlo Method and GPU

高温物理气相合成是合成纳米颗粒的重要方法之一,该过程包含复杂的成核、生长、凝并等颗粒动力学事件,对其进行精准、快速的数值模拟具有重要的指导意义。基于群平衡Monte Carlo方法,本研究采用GPU统一计算设备架构(CUDA)并行算法模拟Ag纳米颗粒的高温物理气相合成过程。为了提高计算效率,颗粒凝并事件考虑了快速Monte Carlo方法。该方法能够很好地描述纳米颗粒的高温物理气相合成过程,与实验测量对比表明,模拟结果基本吻合颗粒群尺度分布,计算速率平均提高13.8%。

Gas synthesis at high temperature is one of the most important methods for producing nanoparticles. It includes complicated particle dynamic events, such as nucleation, growth, coagulation and so on. Hence, accurate and fast numerical simulation is very significant and instructive.Based on Population Balance Monte Carlo(PBMC), parallel computing using Compute Unified Device Architecture(CUDA) under GPU is employed in this research to capture the gas synthesis process of Ag nanoparticles at high temperature. To speed up the calculation, the Fast Monte Carlo method is used to solve the coagulation event. The method mentioned above shows good advantages in gas synthesis process and the computational efficiency is increased by 13.8%. The results are basically in line with the particle size distributions which are attained from experiments.