• 한국전산유체공학회:학술대회논문집
• /
• 2004.03a
• /
• pp.77-82
• /
• 2004

A new anisotropic mesh refinement method is proposed. The new method is based on a simple second order interpolation error indicator. Therefore, it is methodologically direct and intuitive as compared with traditional anisotropic refinement strategies. Moreover, it does not depend on the mesh type. The error indicator is face-wisely calculated for all faces in a mesh and the cell refinement type is determined by the configuration of face markings with a given threshold. For the sake of simplicity, an application for a poisson equation on a triangle mesh is considered. The error field and resultant mesh refinement pattern are compared and effects of the threshold selection are discussed. Applying anisotropic refinement with various thresholds, we observed higher convergence rates than those in the uniform refinement cases.

• Structural Engineering and Mechanics
• /
• v.74 no.4
• /
• pp.547-557
• /
• 2020

In this paper, an automatic adaptive mesh refinement procedure is presented for two-dimensional problems on the basis of a new probabilistic error estimator. First-order perturbation theory is employed to determine the lower and upper bounds of the structural displacements and stresses considering uncertainties in geometric sizes, material properties and loading conditions. A new probabilistic error estimator is proposed to reduce the mesh dependency of the responses dispersion. The suggested error estimator neglects the refinement at the critical points with stress concentration. Therefore, the proposed strategy is combined with the classic adaptive mesh refinement to achieve an optimal mesh refined properly in regions with either high gradients or high dispersion of the responses. Several numerical examples are illustrated to demonstrate the efficiency, accuracy and robustness of the proposed computational algorithm and the results are compared with the classic adaptive mesh refinement strategy described in the literature.

• Structural Engineering and Mechanics
• /
• v.75 no.6
• /
• pp.685-699
• /
• 2020

Polygonal finite element provides a great flexibility in mesh generation of crack propagation problems where the topology of the domain changes significantly. However, the control of the discretization error in such problems is a main concern. In this paper, a polygonal-FEM is presented in modeling of crack propagation problems via an automatic adaptive mesh refinement procedure. The adaptive mesh refinement is accomplished based on the Zienkiewicz-Zhu error estimator in conjunction with a weighted SPR technique. Adaptive mesh refinement is employed in some steps for reduction of the discretization error and not for tracking the crack. In the steps that no adaptive mesh refinement is required, local modifications are applied on the mesh to prevent poor polygonal element shapes. Finally, several numerical examples are analyzed to demonstrate the efficiency, accuracy and robustness of the proposed computational algorithm in crack propagation problems.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2004.10a
• /
• pp.79-82
• /
• 2004

In this study, a new refinement technique for 3-dimensional hexahedral element mesh is proposed, which is aimed at the control of mesh density. With the proposed scheme the mesh is refined adaptively to the elemental error which is estimated by 'a posteriori' error estimator based on the energy norm. A desired accuracy of an analysis i.e. a limit of error defines the new desired mesh density map on the current mesh. To obtain the desired mesh density, the refinement procedure is repeated iteratively until no more elements to be refined exist. In the algorithm, at first the regions of mesh to be refined are defined and, then, the zero-thickness element layers are inserted into the interfaces between the regions. All the meshes in the regions, in which the zero-thickness layers are inserted, are to be regularized in order to improve the shape of the slender elements on the interfaces. This algorithm is tested on a simple shape of 2-d quadrilateral element mesh and 3-d hexahedral element mesh. A numerical example of elastic deformation of a plate with a hole shows the effectiveness of the proposed refinement scheme.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.18 no.11
• /
• pp.2932-2943
• /
• 1994

An efficient approach to the automatic construction of effective quadrilateral finite element meshes for two-dimensional analysis is presented. The procedure is composed of, firstly, an initial mesh generation and, secondly, an h-version of adaptive refinement based on error analysis. As for an initial mesh generation scheme, a modified looping algorithm has been employed. For the adaptive refinement process, an error indicator obtained by computing the residual error of the equilibrium equations in the energy norm with a relaxation factor has been employed. Examples of mesh generation and self-adaptive mesh improvements are given. These example solutions demonstrate that an effective mesh for a given error tolerance can be obtained in a few steps of the analysis processes.

• Journal of the Korean Institute of Telematics and Electronics D
• /
• v.34D no.7
• /
• pp.37-44
• /
• 1997

A dadaptive refinement tetrahedron mesh has been presented for using in full band GaAs monte carlo simulation. A uniform tetrahedron mesh is used without regard to energy values and energy variety in case of the past full band simulation. For the uniform tetrahedron mesh, a fine tetrahedron is demanded for keeping up accuracy of calculation in the low energy region such as .GAMMA.-valley, but calculation time is vast due to usin gthe same tetrahedron in the high energy region. The mesh of this study, thererfore, is consisted of the fine mesh in the low energy and large variable energy region and rough mesh n the high energy. The density of states (DOS) calculated with this mesh is compared with the one of the uniform mesh. The DOS of this mesh is improved th efive times or so in root mean square error and the ten times in the correlation coefficient than the one of a uniform mesh. This refinement mesh, therefore, can be used a sthe basic mesh for the full band GaAs monte carlo simulation.

• International Journal of Aeronautical and Space Sciences
• /
• v.18 no.4
• /
• pp.601-613
• /
• 2017

An adjoint-based error estimation and mesh adaptation study is conducted for two-dimensional viscous flows on unstructured hybrid meshes. The error in an integral output functional of interest is estimated by a dot product of the residual vector and adjoint variable vector. Regions for the mesh to be adapted are selected based on the amount of local error at each nodal point. Triangular cells in the adaptive regions are refined by regular refinement, and quadrangular cells near viscous walls are bisected accordingly. The present procedure is applied to single-element airfoils such as the RAE2822 at a transonic regime and a diamond-shaped airfoil at a supersonic regime. Then the 30P30N multi-element airfoil at a low subsonic regime with a high incidence angle (${\alpha}=21deg.$) is analyzed. The same level of prediction accuracy for lift and drag is achieved with much less mesh points than the uniform mesh refinement approach. The detailed procedure of the adjoint-based mesh refinement for the multi-element airfoil case show that the basic flow features around the airfoil should be resolved so that the adjoint method can accurately estimate an output error.

• Journal of Korean Port Research
• /
• v.10 no.2
• /
• pp.61-68
• /
• 1996

The main problems in structural analysis by Finite Eelement Method are difficulty in making data file and error estimation. For decreasing these problems' pays. have been suggesting the adaptive mesh refinement and error estimation method. Posteriory error estimation methods suggested by Jang[1], Babuska[2,3], Ohtsubo[8,9], and this paper. Comparing these methods and examine their properties. According this paper, In the problem supposed having singularity, the method suggested by this paper is good, But the problem supposed having no singularity, the method suggested by Jang[1] is good. For decreasing the effect of initial mesh in p-refinement, make application h-refinement at first and apply p-refinement, and confine polynomial's degree to two, for making program simply by plural mesh models are not needed.

• Structural Engineering and Mechanics
• /
• v.17 no.3_4
• /
• pp.379-391
• /
• 2004

This paper deals with the development of h-version adaptive mesh refinement and recovery strategy using variable-node elements and its application to various engineering field problems with 2D quadrilateral and 3D hexahedral models. The variable-node elements which have variable mid-side nodes on edges or faces are effectively used in overcoming some problems in connecting the different layer patterns of the transition zone between the refined and coarse mesh. A modified recovery technique of gradients adequate for variable-node elements and proper selection of error norms for each engineering field problems are proposed. In the region in which the error is greater than the permissible refinement error, the mesh is locally refined by subdivision. Reversely, in some parts of the domain having the error smaller than the permissible recovery error, the mesh is locally recovered (coarsened) by combination. Hierarchical structures (e.g. quadtrees and octrees) and element-based storage structures are composed to perform this adaptive process of refinement and recovery. Some numerical examples of a 3D heat conduction analysis of the concrete with hydration heat and a 2D flow analysis of vortex shedding show effectiveness and validity of the proposed scheme.

• Journal of Mechanical Science and Technology
• /
• v.20 no.12
• /
• pp.2189-2196
• /
• 2006

The present work is concerned with the development of a procedure for adaptive computations of shear localization problems. The maximum jump of equivalent strain rates across element boundaries is proposed as a simple error indicator based on interpolation errors, and successfully implemented in the adaptive mesh refinement scheme. The time step is controlled by using a parameter related to the Lipschitz constant, and state variables in target elements for refinements are transferred by $L_2$-projection. Consistent tangent moduli with a proper updating scheme for state variables are used to improve the numerical stability in the formation of shear bands. It is observed that the present adaptive mesh refinement procedure shows an excellent performance in the simulation of shear localization problems.