• Transactions of Materials Processing
• /
• v.5 no.1
• /
• pp.55-60
• /
• 1996

In the finite element analysis of metal forming processes the updated Lagrangian approach has been widely and effectively used to simulate the non-steady state problems. however some difficulties have arisen from abrupt flow change as in extrusion through square dies. In the present work an ALE(arbitrary Lagrangian-Euleria) finite element formulation for deforma-tion analysis are presented fro rigid-viscoplastic materials. The developed finite element program is applied to the isothermal analysis of square die extrusion of a square section. The computational results are compared with those by the updated Lagrangian finite element analysis.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.30 no.10 s.253
• /
• pp.1320-1327
• /
• 2006

Parallel procedure and performance of two-step Eulerian code have not been reported sufficiently yet even though it was developed and utilized widely in the impact simulation. In this study, parallel strategy of two-step Eulerian code was proposed and described in detail. The performance was evaluated in the self-made linux cluster computer. Compared with commercial code, a relatively good performance is achieved. Through the performance evaluation of each computation stage, remap is turned out to be the most time consuming part among the other part such as FE processing, communication, time marching etc.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.2
• /
• pp.208-212
• /
• 2013

Due to the singularity of the within-class scatter, linear discriminant analysis (LDA) becomes ill-posed for small sample size (SSS) problems. An extension of LDA, the null space-based LDA (NLDA) provides good discriminant performances for SSS problems. In this paper, by applying the Lagrange technique, the procedure of transforming the problem of finding the feature extractor of NLDA into a linear equation problem is derived.

• Journal of Korean Institute of Industrial Engineers
• /
• v.22 no.4
• /
• pp.741-751
• /
• 1996

The three dimensional cutting-stock problem is to maximize the total value of pieces which are smaller cubics-cut from a original cubic stock. This paper suggests a method to maximize the total value of different size cut pieces using the orthogonal non-guillotine cut technique. We first formulated a zero-one integer programming, then developed a Lagrangeon relaxation method far the problem. The solutions were given by using a brunch-end-bound technique associates with Lagrangean relaxation, which guarantees an optimal solution.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10a
• /
• pp.39-42
• /
• 2006

뉴럴네트워크에서 학습은 에러를 줄이는 방법으로 구현 된다. 이 때 parameter 공간에서 Risk function은 multi-minima potential로 표현 될 수 있으며 우리의 목적은 global minimum weight 좌표를 얻는 것이다. 이전의 연구로는 Attouch et al.의 damped oscillator 방정식을 이용한 방법이 있고, Qian의 critically damped oscillator를 통한 steepest descent의 momentum과 learning parameter 유도가 있다. 우리는 이 두 연구를 참고로 manifold 상에서 최단 경로인 geodesic을 Newton 역학의 Lagrangian에 적용함으로써 adaptive steepest descent 학습법을 얻었다. 우리는 이 새로운 방법을 Rosenbrock 과 Griewank 포텐셜들에 적용하여 그 성능을 알아 본다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.557-558
• /
• 2022

본 논문에서는 인공신경망을 통해 화면에 투영된 거품입자를 효율적으로 생성할 수 있는 기법에 대해 소개한다. 유체 시뮬레이션 기반으로 바다거품을 계산하기 위해서는 유체역학과 수치해석학에 대한 이해가 필요하며, 유속의 유기물, 풍속 등 다양한 물리적 요소를 고려해야하기 때문에 복잡하고 계산양이 커진다. 오일러리안(Eulerian)접근법에서는 격자의 해상도가 커지게 되고, 라그랑지안(Lagrangian)접근법에서는 입자의 개수가 많아지기 때문에 이 문제를 다루기 쉽지 않은 문제이다. 이러한 문제를 완화하기 위해 본 논문에서는 인공신경망을 이용한 분류 모델 학습을 통해 3차원 유체 시뮬레이션으로부터 투영된 2차원 스크린 이미지로부터 거품이 생성될 위치를 예측한다. 결과적으로 물의 스크린에 투영된 물 입자의 깊이와 가속도로부터 거품의 생성 위치를 예측함으로서 복잡한 수치해석학 없이 학습을 통해 효율적으로 거품을 표현하는 결과를 보여준다.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.24 no.2
• /
• pp.97-108
• /
• 2012

A Lagrangian particle tracking model coupled with the ECOM3D were used to study on the behavior of fresh water released from the Namgang Dam in terms of residence time in Kangjin Bay, South Sea, Korea. Model was calibrated until skill cores for elevation, velocity, temperature and salinity are satisfied over 85%. In the numerical simulation, particles were released in 1 hour time interval from the northern boundary. The different patterns of particle trajectory are identified under the varying dynamics from tidal to density-driven current. The average residence time of total particles are approximately 65.9 hours in the entire Kangjin Bay. The average residence time were increased from 55~65 to 70~80 hours during maximum discharge period. Discharge rate of fresh water and average residence time in the Kangjin Bay is high correlated with correlation coefficient over 0.81.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.27 no.1
• /
• pp.1-8
• /
• 2014

Using the level set and the meshfree methods, we develop a topological shape optimization method applied to linear elasticity problems. Design gradients are computed using an efficient adjoint design sensitivity analysis(DSA) method. The boundaries are represented by an implicit moving boundary(IMB) embedded in the level set function obtainable from the "Hamilton-Jacobi type" equation with the "Up-wind scheme". Then, using the implicit function, explicit boundaries are generated to obtain the response and sensitivity of the structures. Global nodal shape function derived on a basis of the reproducing kernel(RK) method is employed to discretize the displacement field in the governing continuum equation. Thus, the material points can be located everywhere in the continuum domain, which enables to generate the explicit boundaries and leads to a precise design result. The developed method defines a Lagrangian functional for the constrained optimization. It minimizes the compliance, satisfying the constraint of allowable volume through the variations of boundary. During the optimization, the velocity to integrate the Hamilton-Jacobi equation is obtained from the optimality condition for the Lagrangian functional. Compared with the conventional shape optimization method, the developed one can easily represent the topological shape variations.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.8
• /
• pp.8-14
• /
• 2019

Bird-strike is one of the most important design factors for safety in the aviation industry. Bird-strikes have been the cause of significant damage to aircraft and rotorcraft structures and the loss of life. This study used DYTRAN software to simulate the transient response of an Euler-Lagrangian composite helicopter blade that has been impacted by a bird. The Arbitrary Lagrangian Eulerian (ALE) method and a suitable equation of state were applied to model the bird. ALE was applied to the bird-strike analysis due to the large difference between the properties of the blade and bird. The debris of the bird was assumed to be a fluid and applied as Euler elements after the collision. Through the analysis of bird impacts, the leading-edge of the rotor blade (50.8 mm) was used to identify a positive margin of 1.18 based on the TSAI-FILL criteria. The results are assessed to be sufficiently reliable and may be evaluated to replace tests with various analysis conditions. The structural stability of the rotor blade could be assessed by applying various load conditions and different modeling methods in the future.

• Journal of the Korean Society of Urban Environment
• /
• v.18 no.4
• /
• pp.511-521
• /
• 2018

Despite the limitations and difficulty in the application of CALINE3 model for air dispersion prediction of roads and tunnels construction businesses in South Korea, the model is being used in all roads construction projects. This study compared the predicted values of CALINE3 and AERMOD model that is suggested by the US EPA, to the values of GRAL model, a Lagrangian particle tracking model developed in Europe, by applying the models to the existing roads of the urban areas. The result showed low relevance to the actual measurement value in the case of CALINE3 model, thus displaying a low trusted value when applying to the urban areas. In the case of using AERMOD model, the predicted values were overly expressed compared to the actual measurement value, thus leading to the need of adding a No2 conversion method to the model in the future. In the case of GRAL model, a Lagrangian particle tracking model, the relevance between the actual and predicted values were high as the model considers the surrounding topography and the buildings all together, thus confirming that the model can be used for air dispersion prediction of the roads in the urban areas. Lastly, the result of this study testing the air prediction models in Jeongneung Measuring Station points that it is necessary for the future studies to expand the testing areas and test the validity of the models continuously.