탄소성 수치해석상의 과다응력평가에 관한 연구

민옥기,임욱선,박명식 연세대학교 산업기술연구소 1986 논문집 Vol.18 No.1

One of the inherent errors in the numerical analysis of an elasto-plastic problem is the error occurred in the plastic stress computation. If an increment in the load state yield the stress point beyond the yield surface. then a correction of this excess stress must be made in order to meet the yield condition. Tangent Predictor Radial Return Method (TPRRM). Secant Stiffness Method (SSM), and Elastic Predictor Radial Corrector Method (EPRCM) have been studied for a common elasto-plastic condition of von Mises. Errors in stress prediction are expressed with two parameters in the π-plane: (I) the angle between the analytical and the computed final stress points and (II) the difference between the analytical and the computed yield surface radii. Single Step Correction(SSC) and Multi-Step Correction(MSC)are adopted in order to compare the errors of these methods. For single step correction, the results of EPRCM is generally more accurate than the results of SSM or TPRRM. The results of MSC further yield smaller errors in angle and radius and show more stable numerical solutions.

점소성 대변위 유한요소해석에서 시간증분계수의 영향

민옥기,이선표 연세대학교 산업기술연구소 1988 논문집 Vol.20 No.2

In this paper, the accuracy, convergency and efficiency of the time integral schemes which determine the viscoplastic (creep) strain rate occuring in an incremental elasto-viscoplastic finite element analysis are compared and discussed. Forward, central and backward difference schemes are employed in the temporal domain and the isoparametric 2-D quadrilateral elements are used in the spatial domain. The viscoelastic Maxwell material is incorporated with the elasto -plastic idealization in order to model the visco-plastic constitutive relation. A homogeneous function is used in the fluidity parameter. Total Lagrangian formulation is adopted for the larger displacement. For the convergence of the combined non-linearity. Newton-Raphson method is incrementally employed. Internally pressurized thick cylinder and eccentrically loaded column are chosen to test the effect of time increment on the solution.

점용접이음의 유한요소 수식화에 관한 연구

민옥기,김성환,권해성 연세대학교 대학원 1989 延世論叢 Vol.25 No.1

This paper is concerned with the formulation for the finite element analysis of a 3-dimensional structure, which is composed of spot-welded plates. A spot element is introduced to represent a spot-welded point, which corresponds to an elastic connection infinite element. Since the spot element is a modified one of the 3-dimensional beam element, it has 6-d.o.f. at each nodal point. The shell element used here has 6-d.o.f. at each nodal point including a rotational d.o.f. perpendicular to itself, which can be viewed as the decoupled superposition of membrane and bending elements. The spot element connects two plates attached. The numerical results of the spot welder plate models with one or more spot welding points are as follows. (1) Under the transverse loading, the displacements of the spot welded plate models are in good agreement with those of the plate without spot welding. (2) Under the tensile loading, the displacements of the spot welded plate models are in good agreement with those of the plate without spot welding when the array of the spot welding points is the same direction as the loading direction. (3) Under inplane bonding, the displacements of all the spot welded plate models except one spot plate model are in good agreement with the plate without spot welding.

비선형 과도열전달문제의 유한요소해석에 관한 연구

민옥기 연세대학교 산업기술연구소 1984 논문집 Vol.16 No.2

The heat energy generated by the combustion reaction sharply increases the temperature of the material around the combustion zone while some part of the generated energy are dissipated through heat flow. The numerical approximation for the step-like temperature profile across the reaction zone always shows oscillations. The propagation of the combustion front is a function of thermal properties of the material and the generation rate of internal energy. In this study the coal combustion energy is taken as the generation rate of internal energy and the temperature dependent material properties of coal is considered in the numerical formulation. The temperature profiles across the combustion front shows severe undulations with large time increment and large rate of combustion energy. The multiple locations of combustion front is detected under this temperature undulations, since the combustion front is assumed as isotherms of a reaction temperature. It is shown that the propagation velocity of the combustion front can be estimated by numerical simulations.

주행 테이프 부상 특성의 비선형 수식화

민옥기,권해성 연세대학교 산업기술연구소 1991 논문집 Vol.23 No.2

This study analyzes the running mechanism of flexible and thin tape above rotating protuberance through the numerical simulation. The scope of analysis is confined to the phenomena of elasto-hydrodynamic lubrication between the rotating cylinder with a protuberance and the running tape. This model is based on the equation of plate considering the effect of geometric nonlinearity and the modified Reynolds equation. Finite element method is adopted as a numerical simulation technique to solve the above two coupled nonlinear equations. In numerical analysis, the effects of tape tension and the relative velocity are basically evaluated.

유한요소 모델의 행렬식 계수폭의 축소에 관한 연구

민옥기 연세대학교 산업기술연구소 1985 논문집 Vol.17 No.1

A relatively straight-forward simultaneous renumbering algorithm for both element and node numbers in the finite element mesh discretization is presented in this paper. The algorithm is applicable to two and three-dimensional models. The advantages of simultaneous renumbering of node and element tags in the finite element mesh discretization are illustrated. A possible small capacity micro-computer application of the algorithm is discussed, in explaining the algorithm in BASIC.

Mindlin 판의 강성 및 질량행렬 계산법의 수정을 통한 유한 요소의 성능향상

김용우,민옥기,박춘수 연세대학교 산업기술연구소 1990 논문집 Vol.22 No.2

An improved scheme for calculating stiffness and mass matrices of Mindlin plate is suggested through the eigen analysis of rectangular plates. The significant error due to locking phenomenon can be alleviated by employing the modified shape function in calculating stiffness matrix without reduced integration. Further, the application of Gauss-Lobatto quadrature to mass matrix improves the behavior of numerical results in cooperation with the stiffness matrix calculated by the modified shape function. To illustrate these facts, 9-noded rectangular plate element is used, which is widely used in practice. Results for rectangular plate with all edges simply-supported and all edges built-in are presented, which demonstrated the effects of the modified shape function and the validity of Gauss-Lobatto quadrature for mass matrix.

회전헤드장치에서 헤드창을 고려한 박막탄성체의 수치해석

권해성,민옥기 연세대학교 산업기술연구소 1993 논문집 Vol.25 No.1

The running mechanism of flexible and thin foil above rotating protrusion is analyzed throught a numerical simulation. The scope of analysis is confined to the phenomena of elastohydrodynamic lubrication between the rotary drum with a running protrusion and thin foil. This mathematical analysis is based on the modified Reynolds equation and the equation of plate, considering the protrusion geometry and the effect of geometric nonlinearity. Finite element method is adopted as a numerical simulation method to solve the above coupled nonlinear equations. In numerical analysis, the effects of head window in rotary drum are evaluated. And the analysis results of the window model are compared with the those of no window model.

Mindlin판의 중량 최소화에 관한 연구

김세윤,민옥기,김용우 연세대학교 산업기술연구소 1991 논문집 Vol.23 No.2

This paper deals with the minimum weight optimization of Mindlin plate under stress, displacement, eigen value, and design variable constraints. Modified shape functions and a full integration are employed in finite element analysis to remove errors due to spurious constraints and spurious modes. The weight of Mindlin plate, modeled by 9-noded rectangular Mindlin plate element, is minimized by choosing the thickness of element as a design variable and by using gradient projection method with analysis of design sensitivity for constraints. The numerical example is selected to compare the result of the present formulation with that of published one. This comparison shows that the procedure presented in this paper can be applicable to weight minimization of plate in practice.

반대칭 변위조건을 갖는 축대칭 얇은 쉘의 유한요소해석

김성일,이상신,민옥기 연세대학교 산업기술연구소 1987 논문집 Vol.19 No.1

This paper describes the application of the 2-dimensional finite element method (FEM) based on the concept of Direct Stiffness Method (DSM) to the linear elastic analysis of axisymmetric thin shells of revolution under both symmetric and antisymmetric boundary conditions. The shell is idealized as a series of conical frusta, joined at two nodal circles. In the present paper the comparison of both symmetric and antisymmetric rigid body displacement results obtained using a 2-dimensional finite element formulation with those obtained from analytical solution is deminstrated. It is concluded that the conical frustum element can be used in predicting the rigid body motion behavior under both symmetric and antisymmetric and antisymmetrtc initial displacement boundary conditions.