GPU-accelerated visualisation of ADS granular flow target model. (2015)

• Record Type: Journal Article
• Authors: Tian, Yan-Shan; Zhou, Qingguo; Sun, Hong-Yu; Wu, Jiong; Zhang, Xun-Chao; Li, Kuan-Ching
• Is Part Of: International journal of high performance computing and networking. Volume 8:Number 4(2015)
• Page Start: 381
• Page End: 389
• Publication Date: 2015
• Subjects: parallel acceleration -- compute unified device architecture -- CUDA -- open graphics library -- OpenGL -- discrete element method -- DEM -- real-time simulation -- GPU-accelerated visualisation -- particle visualisation -- accelerator-driven sub-critical system -- ADS granular flow target -- geometric modelling -- particle collision detection -- transport simulation -- protons -- neutrons -- nuclear energy -- nuclear waste -- nuclear power -- edge detection -- position estimation -- motion direction -- collision points -- GPUs -- graphical processing units; High performance computing -- Periodicals; Computer networks -- Periodicals; High performance computing; Periodicals; 004.05
• Languages: English
• ISSNs: 1740-0562

Abstract

This paper presents a discrete element method to handle particle collision detection and responses in transport simulation (the simulation of transport of protons and neutrons in granular flow target geometric model) based on GPUs. Discrete element method was adopted in the realisation of large-scale particle visualisation. The method simulates and solves edge detection, position judging, motion direction, calculation of the next collision point using GPU acceleration during the process of transport, and demonstrates the complete interaction process through OpenGL. Results show that the model presented exploits the acceleration of GPUs and has gained remarkable functional improvement compared with traditional method using solely CPUs. In addition, we used the MCNPX to calculate this model with high-speed proton bombardment. The distribution of power energies verifies that the granular flow target model is reliable and feasible.