• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.23 no.5
• /
• pp.475-482
• /
• 2010

In this paper, an efficient two-level parallel domain decomposition algorithm is suggested to solve large-DOF structural problems. Each subdomain is composed of the coarse problem and local problem. In the coarse problem, displacements at coarse nodes are computed by the iterative method that does not need to assemble a stiffness matrix for the whole coarse problem. Then displacements at local nodes are computed by Multi-Frontal Sparse Solver. A parallel version of PCG(Preconditioned Conjugate Gradient Method) is developed to solve the coarse problem iteratively, which minimizes the data communication amount between processors to increase the possible problem DOF size while maintaining the computational efficiency. The test results show that the suggested algorithm provides scalability on computing performance and an efficient approach to solve large-DOF structural problems.

• Computational Structural Engineering
• /
• v.5 no.2
• /
• pp.18-27
• /
• 1992

88년 9월중에 한국과학기술연구원 시스템공학연구소가 당시의 슈퍼컴퓨터 중 최상위 성능을 가진 Cray-2S(4 CPU, 1GB)를 설치함에 따라 국내에도 슈퍼컴퓨터 시대가 열리게 되었으며, 90년 10월에 산업계에서는 최초로 기아자동차에서 Cray Y/MP(1CPU)를 설치한 이래 최근에 국방과학연구소, 삼성그룹에서도 Cray Y/MP계열의 슈퍼컴퓨터를 설치하여 과학기술 계산 및 공학해석에 폭넓게 활용할 전망이다. 따라서 본 고에서는 슈퍼컴퓨터의 정의 및 분류, 특징과 보급현황에 대하여 알아보고 슈퍼컴퓨터 및 병렬처리기술을 이용한 유한요소해석에 관하여 간략히 기술하고저 한다.

• Computational Structural Engineering
• /
• v.11 no.3
• /
• pp.55-70
• /
• 1998

• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.241-242
• /
• 2007

병렬형 마일드 HEV(Hybrid Electric Vehicles)는 동력변환과정이 적어 구동계 전체의 효율이 직렬형에 비해 우수하고 다양한 구조를 가질 수 있으며 기존차량에 적용하기 쉽다는 장점이 있으나 구조 및 제어가 복잡하다. 따라서 병렬형 마일드 HEV의 성능을 예측하고 적절한 제어기를 설계하기 위해서는 구성요소의 종류 및 규격과 제어전략에 따른 HEV의 성능을 해석할 수 있는 체계적인 방법이 필요하다. 따라서 본 논문에서는 Simulink 소프트웨어를 이용한 모듈화 모델링에 의하여 병렬형 HEV의 구성요소를 모델링하고 이로부터 병렬형 HEV의 성능해석 및 운전제어전략의 특성을 비교할 수 있도록 한다.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.20 no.3
• /
• pp.281-286
• /
• 2007

An engineering design problem involving Multi-Disciplinary Analysis(MDA) generally requires a large amounts of computing time for the entire design process, and therefore it is essential to introduce a Multiple Processor System (MPS) for reducing the computing time. However, when applying conventional parallel processing techniques, all of the CAE S/W requited for the MDA should be installed on all the servers making up MPS because of characteristic of MDA and it would be a great expense in CAE S/W licenses. To solve this problem, we propose a Weight-based Multiqueue Load Balancing algorithm for a heterogeneous MPS where performance of servers and CAE S/W installed on each server are different of each other. To validate the performance, a Computational experiments comparing the First Come First Serve algorithm and our proposed algorithm was accomplished.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.23 no.4
• /
• pp.361-368
• /
• 2010

In this paper, the parallel algorithm using MPI(Message-Passing Interface) library is introduced in order to improve numerical efficiency for the staggered method for nearly incompressible and impermeable porous media which was introduced by Park and Tak(2010). The porous media theory and the staggered method are also briefly introduced in this paper. Moreover, we account for MPI library for blocking, non-blocking, and collective communication, and propose combined the staggered method with the blocking and nonblocking MPI library. And then, we present how to allocate CPUs on the staggered method and the MPI library, which is related with the numerical efficiency in order to solve unknown variables on nearly incompressible and impermeable porous media. Finally, the results comparing serial solution with parallel solution are verified by 2 dimensional saturated porous model according to the number of FEM meshes.

• Computational Structural Engineering
• /
• v.12 no.3
• /
• pp.62-67
• /
• 1999

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.33 no.3
• /
• pp.171-178
• /
• 2020

In this study, we develop MPI-OpenMP hybrid parallelization for multibody peridynamic simulations. Peridynamics is suitable for analyzing complicated dynamic fractures and various discontinuities. However, compared with a conventional finite element method, nonlocal interactions in peridynamics cost more time and memory. In multibody peridynamic analysis, the costs increase due to the additional interactions that occur when computing the nonlocal contact and ghost interlayer models between adjacent bodies. The costs become excessive when further refinement and smaller time steps are required in cases of high-velocity impact fracturing or similar instances. Thus, high computational efficiency and performance can be achieved by parallelization and optimization of multibody peridynamic simulations. The analytical code is developed using an Intel Fortran MPI compiler and OpenMP in NURION of the KISTI HPC center and parallelized through MPI-OpenMP hybrid parallelization. Further parallelization is conducted by hybridizing with OpenMP threads in each MPI process. We also try to minimize communication operations by model-based decomposition of MPI processes. The numerical results for the impact fracturing of multiple bodies show that the computing performance improves significantly with MPI-OpenMP hybrid parallelization.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.40 no.11
• /
• pp.972-978
• /
• 2012

This paper discusses the parallelization of multifrontal solution method, widely used for finite element structural analyses, for a shared memory architecture. Multifrontal method is easier than other linear solution methods because the solution procedure implies that unknowns can be eliminated simultaneously. Two innovative ideas are introduced to achieve optimal solver performance on a shared memory computer. Those are pairing two frontal matrices and splitting the frontal matrix in order to reduce the temporal memory space required by independent computing tasks. Performance comparisons between original algorithm and proposed one prove that proposed method is more computationally efficient on current multicore machines.

• Computational Structural Engineering
• /
• v.11 no.3
• /
• pp.78-84
• /
• 1998