Particle-in-cell (PIC) method has got much benefits from GPU-accelerated heterogeneous systems.However,the performance of PIC is constrained by the interpolation operations in the weighting process on GPU (graphic pro...Particle-in-cell (PIC) method has got much benefits from GPU-accelerated heterogeneous systems.However,the performance of PIC is constrained by the interpolation operations in the weighting process on GPU (graphic processing unit).Aiming at this problem,a fast weighting method for PIC simulation on GPU-accelerated systems was proposed to avoid the atomic memory operations during the weighting process.The method was implemented by taking advantage of GPU's thread synchronization mechanism and dividing the problem space properly.Moreover,software managed shared memory on the GPU was employed to buffer the intermediate data.The experimental results show that the method achieves speedups up to 3.5 times compared to previous works,and runs 20.08 times faster on one NVIDIA Tesla M2090 GPU compared to a single core of Intel Xeon X5670 CPU.展开更多
Electrical tomography is, in certain cases, the most attractive method for real imaging of industrial processes, because of its inherent simplicity, rugged construction of the tomographer and high-speed capability. Th...Electrical tomography is, in certain cases, the most attractive method for real imaging of industrial processes, because of its inherent simplicity, rugged construction of the tomographer and high-speed capability. This paper presents examples illus- trating applications of electrical tomography for imaging fluidized beds, bubble columns and pneumatic conveyors. Electrical tomography opens up new ways for processing, imaging and modelling multi-phase flows as shown by 2D and 3D images illus- trating the various types of flow morphology.展开更多
This paper presents an optimization of shadow volume algorithm, which allow a rendering in real-time. This technique is based on previous works which makes it possible to obtain shadows in real-time, although the calc...This paper presents an optimization of shadow volume algorithm, which allow a rendering in real-time. This technique is based on previous works which makes it possible to obtain shadows in real-time, although the calculation of the silhouette requires a pretreatment of the geometry implemented on the CPU （Central Processing Unit）. By using last version of the GPU （Graphic Processing Unit）, the authors propose to implement the calculation of the silhouette on the GPU by using Geometry Shader. The authors present the step which made it possible to lead to a concrete implementation of this algorithm, the modifications which were made, as well as a comparative study of results, followed by a discussion of these results and choices of implementation.展开更多
A novel algorithm to voxelize 3D mesh models with gray levels is presented in this paper.The key innovation of our method is to decide the gray level of a voxel according to the total area of all surfaces contained by...A novel algorithm to voxelize 3D mesh models with gray levels is presented in this paper.The key innovation of our method is to decide the gray level of a voxel according to the total area of all surfaces contained by it.During the preprocessing stage,a set of voxels in the extended bounding box of each triangle is established.Then we travel each triangle and compute the areas between it and its set of voxels one by one.Finally,each voxel is arranged a discrete gray level from 0 to 255.Experiments show that our algorithm gets a comparatively perfect result compared with the prevenient ones and approaches the original models in a more accurate way.展开更多
In order to clarify effects of prior pancaked austenitic structure on microstructure and mechanical properties of transformed martensite in ausformed steel,a super-thin pancaked austenite was processed by multi-pass r...In order to clarify effects of prior pancaked austenitic structure on microstructure and mechanical properties of transformed martensite in ausformed steel,a super-thin pancaked austenite was processed by multi-pass rolling in a 0.03-2.6Mn0.06Nb-0.01Ti(wt%) low alloy steel.The evolution of prior pancaked austenite grain during multi-pass rolling was studied using Ni-30Fe model alloy.Related with the structure and texture in the prior super-thin pancaked austenite in Ni-30Fe alloy,the texture and anisotropy of mechanical properties of transformed martensite in the studied ausformed steel were focused on.There were mainly three kinds of rolling texture components in the super-thin pancaked austenite:Goss {110} 001,copper {112} 111 and brass {110} 112.They were further transformed into the weak {001} 110 and strong {112} 110,{111} 112 texture components in the martensitic structure.The orientation relationship(OR) of lath martensite transformation from pancaked austenite in the ausformed steel deviated larger from the exact Kurdjumov-Sachs(K-S) OR than in the case of equiaxed austenite without deformation.The tensile and yield strengths of the ausformed martensitic steel first decreased and then increased as the angle between tension direction and rolling direction increased.The main reason for the anisotropy of strength was considered as the texture component {112} 110 in martensite.However,the anisotropy of impact toughness was more complex and the main reasons for it are unknown.展开更多
基金Projects(61170049,60903044)supported by National Natural Science Foundation of ChinaProject(2012AA010903)supported by National High Technology Research and Development Program of China
文摘Particle-in-cell (PIC) method has got much benefits from GPU-accelerated heterogeneous systems.However,the performance of PIC is constrained by the interpolation operations in the weighting process on GPU (graphic processing unit).Aiming at this problem,a fast weighting method for PIC simulation on GPU-accelerated systems was proposed to avoid the atomic memory operations during the weighting process.The method was implemented by taking advantage of GPU's thread synchronization mechanism and dividing the problem space properly.Moreover,software managed shared memory on the GPU was employed to buffer the intermediate data.The experimental results show that the method achieves speedups up to 3.5 times compared to previous works,and runs 20.08 times faster on one NVIDIA Tesla M2090 GPU compared to a single core of Intel Xeon X5670 CPU.
文摘Electrical tomography is, in certain cases, the most attractive method for real imaging of industrial processes, because of its inherent simplicity, rugged construction of the tomographer and high-speed capability. This paper presents examples illus- trating applications of electrical tomography for imaging fluidized beds, bubble columns and pneumatic conveyors. Electrical tomography opens up new ways for processing, imaging and modelling multi-phase flows as shown by 2D and 3D images illus- trating the various types of flow morphology.
文摘This paper presents an optimization of shadow volume algorithm, which allow a rendering in real-time. This technique is based on previous works which makes it possible to obtain shadows in real-time, although the calculation of the silhouette requires a pretreatment of the geometry implemented on the CPU （Central Processing Unit）. By using last version of the GPU （Graphic Processing Unit）, the authors propose to implement the calculation of the silhouette on the GPU by using Geometry Shader. The authors present the step which made it possible to lead to a concrete implementation of this algorithm, the modifications which were made, as well as a comparative study of results, followed by a discussion of these results and choices of implementation.
基金the National Natural Science Foundation of China (60903111)
文摘A novel algorithm to voxelize 3D mesh models with gray levels is presented in this paper.The key innovation of our method is to decide the gray level of a voxel according to the total area of all surfaces contained by it.During the preprocessing stage,a set of voxels in the extended bounding box of each triangle is established.Then we travel each triangle and compute the areas between it and its set of voxels one by one.Finally,each voxel is arranged a discrete gray level from 0 to 255.Experiments show that our algorithm gets a comparatively perfect result compared with the prevenient ones and approaches the original models in a more accurate way.
基金supported by the National Basic Research Program of China("973" Program) (Grant No. 2010CB630805)the National Natural Science Foundation of China (Grant No. 51071089 and 51171087)
文摘In order to clarify effects of prior pancaked austenitic structure on microstructure and mechanical properties of transformed martensite in ausformed steel,a super-thin pancaked austenite was processed by multi-pass rolling in a 0.03-2.6Mn0.06Nb-0.01Ti(wt%) low alloy steel.The evolution of prior pancaked austenite grain during multi-pass rolling was studied using Ni-30Fe model alloy.Related with the structure and texture in the prior super-thin pancaked austenite in Ni-30Fe alloy,the texture and anisotropy of mechanical properties of transformed martensite in the studied ausformed steel were focused on.There were mainly three kinds of rolling texture components in the super-thin pancaked austenite:Goss {110} 001,copper {112} 111 and brass {110} 112.They were further transformed into the weak {001} 110 and strong {112} 110,{111} 112 texture components in the martensitic structure.The orientation relationship(OR) of lath martensite transformation from pancaked austenite in the ausformed steel deviated larger from the exact Kurdjumov-Sachs(K-S) OR than in the case of equiaxed austenite without deformation.The tensile and yield strengths of the ausformed martensitic steel first decreased and then increased as the angle between tension direction and rolling direction increased.The main reason for the anisotropy of strength was considered as the texture component {112} 110 in martensite.However,the anisotropy of impact toughness was more complex and the main reasons for it are unknown.
