Shanghai Supercomputer Center

Shanghai, China

Shanghai Supercomputer Center

Shanghai, China
Time filter
Source Type

Miao X.,Shanghai JiaoTong University | Jin X.,Shanghai JiaoTong University | Ding J.,Shanghai Supercomputer Center
Nongye Jixie Xuebao/Transactions of the Chinese Society for Agricultural Machinery | Year: 2015

An efficient condensation parallel computing method for finite element structural analysis was proposed based on the sparse storage techniques and direct sparse solvers. In the proposed method, the process of condensation was converted to the process of solving a series of linear equations, and then the linear equations were solved with a direct sparse solver. It can avoid the storage and computation of many zero elements within the bandwidth in the traditional parallel computing method with variable bandwidth format condensation. Therefore, the memory space can be greatly saved and the amount of computation can be effectively reduced. Finally, the experiment of the finite element numerical simulation for an engine crankshaft was used to validate the proposed method. Test results showed that, compared with the conventional parallel computing method with variable bandwidth format condensations, the proposed parallel computing approach with sparse storage format condensation could considerably save memory space and significantly improve computational efficiency. The larger the size of each subdomain, the effect of the proposed method on aspects of saving memory space and improving computational efficiency was more obvious. The proposed method can be applied to many industrial areas such as aerospace, automobile, energy, civil and architecture to significantly improve the efficiency of engineering design and analysis. ©, 2015, Chinese Society of Agricultural Machinery. All right reserved.

Miao X.,Shanghai JiaoTong University | Jin X.,Shanghai JiaoTong University | Ding J.,Shanghai Supercomputer Center
International Journal of High Performance Computing Applications | Year: 2015

In order to improve the parallel efficiency of large-scale structural dynamic analysis, a hierarchical approach adapted to the hardware topology of multi-core clusters is proposed. The hierarchical approach is constructed based on the strategies of two-level partitioning and two-level condensation. The data for parallel computing is first prepared through two-level partitioning to guarantee the load balancing within and across nodes. Then during the analysis of each time step, the convergence rate of interface problem is significantly improved by further reducing its size with two-level condensation. Furthermore, the communication overheads are considerably reduced by separating the intra-node and inter-node communications and minimizing the inter-node communication. Numerical experiments conducted on Dawning-5000A supercomputer indicate that the hierarchical approach was superior in performance compared with the conventional Newmark algorithm based on the domain decomposition method. © SAGE Publications.

Cao Y.,Shanghai JiaoTong University | Wang P.,Shanghai Supercomputer Center | Jin X.,Shanghai JiaoTong University | Wang J.,Shanghai JiaoTong University | Yang Y.,Shanghai JiaoTong University
Tunnelling and Underground Space Technology | Year: 2012

Structure analysis of the long tunnel is difficult due to the lack of available computing power. Water hammer simulation in the water conveyance tunnel is also complicated because of strong fluid structure interactions (FSIs). In this paper, the multi-scale modeling method is used to simulate water hammer impacts in the long tunnel. The method can not only yield water hammer simulations along the full tunnel length, but also the detailed structural responses of the segment linings. In the proposed partitioned approach, the structural field is solved with the finite-element program LS-DYNA. The fluid field is solved with the CFD software package FLUENT. The interaction between two physical fields is realized using ALE description. A practical case study is presented and the results are discussed in detail. The results provide us with a better understanding of water hammers and their effects on tunnel linings. © 2011 Elsevier Ltd.

Yuan C.,Shanghai JiaoTong University | Puyong W.,Shanghai Supercomputer Center | Xianlong J.,Shanghai JiaoTong University
Journal of Wind Engineering and Industrial Aerodynamics | Year: 2010

In the case of a flexible container subjected to the wind loads, the container has a high susceptibility to flow-induced deformations and vibrations. This paper presents the work of numerical simulations about its dynamic responses under horizontal wind actions. The container body is assumed to be a Belytschko-Tsay membrane, which includes nonlinearities. Water movement and sloshing are modelled using multi-material ALE method. Interaction between the container and fluid is studied by penalty method, and the large eddy simulation (LES) method is used for simulating turbulence. This paper analyses the problems in container stability, deformation and stress distribution. The results indicate that the reinforcement belt cannot only reduce the container stress but also improve container stability in the wind domain. The maximum stress positions are on the lee side corner of the container. Variation in container vertical acceleration is much larger than the horizontal one. © 2010 Elsevier Ltd.

Miao X.,Shanghai JiaoTong University | Jin X.,Shanghai JiaoTong University | Ding J.,Shanghai Supercomputer Center
Concurrency Computation | Year: 2015

With the advancement of new processor and memory architectures, supercomputers of multicore and multinode architectures have become general tools for large-scale engineering and scientific simulations. However, the nonuniform latencies between intranode and internode communications on these machines introduce new challenges that need to be addressed in order to achieve optimal performance. In this paper, a novel hybrid solver that is especially designed for supercomputers of multicore and multinode architectures is proposed. The new hybrid solver is characterized by its two-level parallel computing approach on the basis of the strategies of two-level partitioning and two-level condensation. It distinguishes intranode and internode communications to minimize the communication overheads. Moreover, it further reduces the size of interface equation system to improve its convergence rate. Three numerical experiments of structural linear static analysis were conducted on DAWNING-5000A supercomputer to demonstrate the validity and efficiency of the proposed method. Test results show that the proposed approach was superior in performance compared with the conventional Schur complement method. © 2014 John Wiley & Sons, Ltd.

Ding J.-H.,Shanghai Supercomputer Center | Xian K.-C.,Aerospace System Engineering Shanghai | Han X.,Shanghai Supercomputer Center | Peng F.-J.,Aerospace System Engineering Shanghai | Deng X.-B.,Shanghai Supercomputer Center
Yuhang Xuebao/Journal of Astronautics | Year: 2011

Aiming at the design of space lenticular boom deployable mechanism, an explicit dynamic algorithm combined with high-performance computing resource is adopted to perform non-linear quasi-static simulation of stowing and deploying process. To meet the large-scale calculating demand due to simulation debugging and multiple loadcases, the explicit calculation parallel speed-up strategy is appropriately established by increasing loading rate after comparison and analysis of time cost, stress and energy curves. The results shows that dynamic simulation method can be used to predict the function design, structure reform, and parameter adjustment as well as provide guidance and reference for experiments.

Feng J.,Anhui University of Science and Technology | Sun X.,Anhui University of Science and Technology | Wu C.,Anhui University of Science and Technology | Peng L.,Anhui University of Science and Technology | And 4 more authors.
Journal of the American Chemical Society | Year: 2011

With the rapid development of portable electronics, such as e-paper and other flexible devices, practical power sources with ultrathin geometries become an important prerequisite, in which supercapacitors with in-plane configurations are recently emerging as a favorable and competitive candidate. As is known, electrode materials with two-dimensional (2D) permeable channels, high-conductivity structural scaffolds, and high specific surface areas are the indispensible requirements for the development of in-plane supercapacitors with superior performance, while it is difficult for the presently available inorganic materials to make the best in all aspects. In this sense, vanadium disulfide (VS2) presents an ideal material platform due to its synergic properties of metallic nature and exfoliative characteristic brought by the conducting S-V-S layers stacked up by weak van der Waals interlayer interactions, offering great potential as high-performance in-plane supercapacitor electrodes. Herein, we developed a unique ammonia-assisted strategy to exfoliate bulk VS2 flakes into ultrathin VS2 nanosheets stacked with less than five S-V-S single layers, representing a brand new two-dimensional material having metallic behavior aside from graphene. Moreover, highly conductive VS2 thin films were successfully assembled for constructing the electrodes of in-plane supercapacitors. As is expected, a specific capacitance of 4760 μF/cm2 was realized here in a 150 nm in-plane configuration, of which no obvious degradation was observed even after 1000 charge/discharge cycles, offering as a new in-plane supercapacitor with high performance based on quasi-two-dimensional materials. © 2011 American Chemical Society.

Xie J.,Anhui University of Science and Technology | Wu C.,Anhui University of Science and Technology | Hu S.,Shanghai Supercomputer Center | Dai J.,Anhui University of Science and Technology | And 4 more authors.
Physical Chemistry Chemical Physics | Year: 2012

Modulating the interaction between small gas molecules and solid host materials is becoming increasingly important for the future society due to the alternative energy resources especially for the hydrogen energy. As is known, two catalogues of materials such as two-dimensional (2D) lamellar cavity structures and three-dimensional (3D) infinite tunnel structures have received intensive considerations during the past decades. Herein, we put forward a new alternative that the solid materials with synergic effects of grain-boundary-rich (GBR) structure and 3D hierarchical hollow structure would also be a promising candidate for modulating the gas molecules in solid adsorbents. As expected, our constructed novel 3D hollow hierarchitectures with GBR shells standing on the hollow spherical cavity indeed resulted in the enhanced hydrogen adsorption behavior. The as-prepared 3D hollow hierarchitectures were very uniform in large scale, and the very simple reaction process offers high convenience, short reaction time, and no need for any complex manipulations or equipments. The hollow outlook of the rutile VO 2(R) 3D hierarchitectures is the reminiscence of the hollow cavity of nsutite-type VO 2, while the formation of the VO 2(R) GBR structure is attributed to volume shrink from a unique intergrowth structure of nsutite-type VO 2. The novel gas modulation model with the synergic effect of GBR structure and hierarchical hollow structure may pave a new way for developing materials in energy and environmental fields in the near future. © 2012 the Owner Societies.

He P.,Anhui University of Science and Technology | Li X.,Anhui University of Science and Technology | Kou D.,Shanghai Supercomputer Center | Deng M.,Anhui University of Science and Technology | Liang H.,Anhui University of Science and Technology
Journal of Chemical Physics | Year: 2010

The self-assembled microstructures of amphiphilic block copolymers depend on the selectivity of solvents for each block. By changing the selectivity of solvents, defined in terms of the repulsive interactions between the solvent and the hydrophilic/hydrophobic particles, an extensive simulation study on the spontaneous formation of complex micelles from amphiphilic triblock copolymers in a dilute solution is presented. The dynamic pathways in the formation of these assemblies have been investigated using a particle-based dissipative particle dynamics approach. In addition, the potential mechanism behind the formation of these microstructures has also been studied, which may be helpful in explaining how these aggregates are formed and in understanding the general principle of amphiphilic molecules. © 2010 American Institute of Physics.

Ding J.-H.,Shanghai Supercomputer Center | Jin Y.-L.,Ship Research Institute | Wang H.,Shanghai Supercomputer Center
Chuan Bo Li Xue/Journal of Ship Mechanics | Year: 2015

Ore fines cargo with high moisture content is liable to liquefaction, making free surface emerging and intensifying sloshing behavior, and finally pose a threat to marine transportation security. With the benefit of Arbitrary Lagrangian-Eulerian (ALE) finite element method, detailed modeling and refined simulation were implemented to deal with shipping liquefied ore fines sloshing problem, and to investigate three-dimensional sloshing phenomena and characteristics with certain charging ratio and motion state. Two different numerical nodes were used to make comparison and analysis for rationality and accuracy of results. High-performance computing (HPC) resource was utilized to meet the challenge of large scale computing requirement due to time-consuming solving and load cases. In view of model size and fluid-structure intersection property, two domain decomposition plans were proposed to better improve load balance for multicore processors, and compared with parallel performance data to determine more reasonable parallel acceleration strategy. ©, 2015, China Ship Scientific Research Center. All right reserved.

Loading Shanghai Supercomputer Center collaborators
Loading Shanghai Supercomputer Center collaborators