Font Size: a A A

Research On High Performence Computation In The Discrete Element Analysis Of Vibration Screen Of A Screen Cleaning Car

Posted on:2015-02-14Degree:MasterType:Thesis
Country:ChinaCandidate:Z H ChenFull Text:PDF
GTID:2530304892480934Subject:(degree of mechanical engineering)
Abstract/Summary:
The screening process of the screen cleaning Car’s vibrating screen is a complicated stochastic process.Since 1979,some scholars began to make the research of the material screening according to the discrete element method,due to the calculation capacity for the movement of the particle which according to the discrete element method is very big,it need to take the solution with the aid of computer.What is the main content of this article research is that how to make use of parallel computing in virtual scene simulation of particle collision and how you can make it vivid display,and how to accomplish the high performance computing.In recent years,with the development of manufacturing technology,GPU has gained good development;after the launch of NVIDIA’s CUDA,there are great developments on GPU’s general calculation based on CUD A technology.Because CUDA C is simple and easy to understand,understanding how to carry out parallel computing on GPU is quite easy for developers.In this paper CUDA C language is written for the small ball collision detection,and OpenGL forrendering is also used,to display the collision detection graphical and make the collision detection process simulation more intuitive for people to understand.In this paper the collision detection of serial is compared with parallel computing,and C++and OpenGL language are used for serial program writing to bring about small balls collision simulation;later,CUDA-C is used to make parallel operation for generating small balls’ data and forrendering is done by interoperating with OpenGL,and finally small balls collision simulation is realized.The experimental results show that the graphical display can make developer’s thoughts more intuitive for people to understand,and the results of using parallel programs written in CUDA C is much better than that written in OpenGL and C++,and parallel programs using GPU can improve speed and performance to a great amount.
Keywords/Search Tags:CUDA, OpenGL, Parallel Computing, Visualization
Related items