• Structural Engineering and Mechanics
• /
• v.8 no.1
• /
• pp.73-84
• /
• 1999

This paper is concerned with shape optimization problems by the boundary element method (BEM) emphasizing the use of a reduced basis reanalysis technique proposed recently by the author. Problems of this class are conventionally carried out iteratively through an optimizer; a sequential quadratic programming-based optimizer is used in this study. The iterative process produces a succession of intermediate designs. Repeated analyses for the systems associated with these intermediate designs using an exact approach such as the LU decomposition method are time consuming if the order of the systems is large. The newly developed reanalysis technique devised for boundary element systems is utilized to enhance the computational efficiency in the repeated system solvings. Presented numerical examples on optimal shape design problems in electric potential distribution and elasticity show that the new reanalysis technique is capable of speeding up the design process without sacrificing the accuracy of the optimal solutions.

• Structural Engineering and Mechanics
• /
• v.8 no.5
• /
• pp.439-452
• /
• 1999

A simple numerical scheme suitable for tracing the fracture propagation path for structures idealized by means of Hillerborg's classical cohesive crack model is presented. A direct collocation, multidomain boundary element method is adopted for the required space discretization. The algorithm proposed is necessarily iterative in nature since the crack itinerary is a priori unknown. The fracture process is assumed to be governed by a path-dependent generally nonlinear softening law. The potentialities of the method are illustrated through two examples.

• Proceedings of the IEEK Conference
• /
• 2001.06b
• /
• pp.329-332
• /
• 2001

This paper proposes a capacitance extraction algorithm based on boundary element method and describes the implemented 2-dimension extractor based on the proposed algorithm. The proposed algorithm uses a generalized conjugate residual iterative algorithm with a hierarchical subdivision. The implemented 2-D extractor computes the capacitances of complicated 2-D geometry of ideal conductors in uniform dielectric and can be efficiently used in the VLSI layout designs due to its user-friendly GUI.

• Journal of Ship and Ocean Technology
• /
• v.4 no.2
• /
• pp.13-23
• /
• 2000

A low-order potential based boundary element method is applied to the prediction of the flow around the cavitating propeller in steady or in unsteady inflow. For given cavitation number, the cavity shape is determined in an iterative manner until the kinematic and the dynamic boundary conditions are both satisfied on the approximate cavity boundary. In order to improve the solution behavior near the tip region, a hyperboloidal panel geometry and a modified split panel method are applied. The method is then extended to include the analysis of time-varying cavitating flows around the propeller blades via a time-step algorithm in time domain. In the method, the steady state oscillatory solution is obtained by incremental stepping in the itme domain. Finally, the present method is validated through comparison with other numerical results and experimental data.

• Journal of the Society of Naval Architects of Korea
• /
• v.34 no.4
• /
• pp.53-60
• /
• 1997

Nonlinear free surface flow phenomena have teen studied by several kinds of numerical methods, of which boundary element method has been known as most promising one. There, however, remain many difficulties to be solved in the numerical procedures by boundary element analysis. In this paper, an efficient calculation of elemental integrals and iterative solution algorithm for the resulting system of equations were thoroughly investigated in order to enhance the procedure of the boundary element analysis. Advantages of the herein developed boundary element analysis code are demonstrated in terms of accuracy and convergence for typical boundary-value problems with free surface.

• Proceedings of the KIEE Conference
• /
• 1999.07a
• /
• pp.130-132
• /
• 1999

The finite element method (FEM) is suitable for the analysis of a complicated region that includes nonlinear materials, whereas the boundary element method (BEM) is naturally effective for analyzing a very large region with linear characteristics. Therefore, considering the advantages in both methods, a novel algorithm for the alternate application of the FEM and BEM to magnetic field problems with the open boundary is presented. This approach avoids the disadvantages of the typical numerical methods with the open boundary problem such as a great number of unknown values for the FEM and non-symmetric matrix for the Hybrid FE-BE method. The solution of the overall problems is obtained by iterative calculations accompanied with the new acceleration method.

• Journal of Mechanical Science and Technology
• /
• v.18 no.3
• /
• pp.460-470
• /
• 2004

A computational analysis of engineering problems with moving domain or/and boundary according to either Lagrangian or Eulerian approach may encounter inherent numerical difficulties, the extreme mesh distortion in the former and the material boundary indistinctness in the latter. In order to overcome such defects in classical numerical approaches, the ALE(arbitrary Lagrangian Eulerian) method is widely being adopted in which the finite element mesh moves with arbitrary velocity. This paper is concerned with the ALE finite element formulation, aiming at the dynamic response analysis of baffled fuel-storage container in yawing motion, for which the coupled time integration scheme, the remeshing and smoothing algorithm and the mesh velocity determination are addressed. Numerical simulation illustrating theoretical works is also presented.

• Proceedings of the KIEE Conference
• /
• 2000.07e
• /
• pp.99-103
• /
• 2000

The finite element method (FEM) is suitable for the analysis of a complicated region that includes nonlinear materials, whereas the boundary element method (BEM) is naturally effective for analyzing a very large region with linear characteristics. Therefore, considering the advantages in both methods, a novel algorithm for the alternate application of the FEM and BEM to magnetic field problems with the open boundary is presented. This approach avoids the disadvantages of the typical numerical methods with the open boundary problem such as a great number of unknown values for the FEM and non-symmetric matrix for the Hybrid FE-BE method. The solution of the overall problems is obtained by iterative calculations accompanied with the new acceleration method.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.348-352
• /
• 2004

In finite element analysis of mechanical behavior of weld, typical process is first to obtain a finite element model containing residual stress by conducting welding analysis and then to examine the computational specimen for various external loading. The numerical specimen with residual stress has irregular boundary lines since one usually begins the welding analysis from a body having regular straight boundary lines and large thermal contraction takes place during cooling of weld metal. We notice that these numerical weld specimens are different from the real weld specimens as the real specimens are usually cut from a bigger weld part and consequently have straight boundaries neglecting elastic relaxation associated with the cutting. In this paper, an iterative finite element method is described to obtain a weld specimen which is bounded by straight lines. The stress distributions of two types of weld specimen, one with regular and the other with irregular boundaries, are compared to check the effect of the boundary shape. Results show that the stress distribution can be different when large plastic deformation is induced by the application of external loading. In case of elastic small deformation, the difference turns out almost negligible.

• Journal of applied mathematics & informatics
• /
• v.8 no.3
• /
• pp.935-942
• /
• 2001

A finite element approximation of a fourth-order nonlinear boundary value problem is given. In the direct implementation, a nonlinear system will be obtained and also a full size matrix will be introduced when Newton’s method is adopted to solve the system. To avoid this difficulty we introduce an iterative scheme which can be shown to converge the positive solution of the system lying between 0 and $sin{\pi}x$.