• Computers and Concrete
• /
• 제23권4호
• /
• pp.235-243
• /
• 2019

The discontinuous Galerkin method (DGM) has become widely used as it possesses several qualities, such as a natural ability to dealing with discontinuities. DGM has its major success related to fluid mechanics. Its major importance is the ability to deal with discontinuities and still provide high order of approximation. That is an important advantage when simulating cracking propagation. No remeshing is necessary during the propagation, since the crack path follows the interface of elements. However, DGM comes with the drawback of an increased number of degrees of freedom when compared to the classical continuous finite element method. Thus, it seems a natural approach to combine them in the same simulation obtaining the advantages of both methods. This paper proposes the application of the combined continuous-discontinuous Galerkin method (CDGM) to crack propagation. An important engineering problem is the simulation of crack propagation in concrete structures. The problem is characterized by discontinuities that evolve throughout the domain. Crack propagation is simulated using CDGM. Discontinuous elements are placed in regions with discontinuities and continuous elements elsewhere. The cohesive zone model describes the fracture process zone where softening effects are expressed by cohesive zones in the interface of elements. Two numerical examples demonstrate the capacities of CDGM. In the first example, a plain concrete beam is submitted to a three-point bending test. Numerical results are compared to experimental data from the literature. The second example deals with a full-scale ground slab, comparing the CDGM results to numerical and experimental data from the literature.

• 한국지반공학회논문집
• /
• 제19권2호
• /
• pp.247-254
• /
• 2003

최근에 건설공사에서 발생하는 환경문제에 대한 관심이 고조되는 가운데 저소음, 저진동 공법인 매입말뚝공법이 건설현장에 많이 적용되고 있다. 이러한 건설여건의 변화에 부흥하여 국내에서는 지반을 굴착하고 시멘트 페이스트를 주입한 후 기성말뚝을 삽입하는 SIP공법(Soil-Cement Injected Precast Pile)의 적용이 늘고 있으나, 현재 국내 지반조건에 맞는 고유의 관계식이 없는 실정이다. 이와 같은 문제해결에 기여하기 위해, 본 연구에서는 SIP말뚝 주면과 지반사이의 저항력을 직접전단시험으로 측정하여 그 경향을 분석하였다. 직접전단시험 결과 SIP말뚝의 주면 저항력 성분이 수직하중에 가장 크게 영향을 받는다는 것을 밝혀내고 국내 주요 토질인 SM, SC 토질에서 이를 인자로 한 최대주면저항력 산정식을 제안하였다. 또한 말뚝 표면과 주면 지반과의 전단저항 거동 특성을 수치적으로 정확히 묘사하기 위해 Duncan(1971)의 쌍곡선 모델식과 새로운 비선형 수치모델식에 따른 계수산정식을 제안하였다.

• 한국지능시스템학회논문지
• /
• 제22권6호
• /
• pp.774-779
• /
• 2012

본 논문은 brain-computer interface (BCI)를 통해 움직임 상상 시 측정된 뇌-활동전위신호(EEG)에 내포된 행동의도의 패턴을 보다 정확하게 분류하기 위한 최적 EEG 채널 선택 기법을 제안한다. 기존의 EEG 측정실험에서는 실험 설계자에 의해 대뇌 기능적 피질 분류를 이용하여 인위적으로 선별된 채널을 활용하거나 측정기기가 수용 가능한 전체 채널을 사용해왔으며, 일정 수준의 패턴분류 정확도를 얻을 수 있었지만 다수의 채널로 인해 Common Spatial Pattern (CSP) 등의 패턴특징 추출 시 overfit 및 계산 복잡도 증가의 문제가 발생되었다. 이를 극복하기 위하여 방안으로 본 논문에서는 binary particle swarm optimization (BPSO)을 기반으로 다수의 채널 중 최적 채널을 자동으로 선택하고, 각각의 채널에 대한 impact factor를 부여함으로써 중요 채널 부근의 채널들에 가중치를 부여하는 선택방법을 제안하였으며, Support Vector Machine (SVM)을 이용하여 다수의 채널을 사용 하였을 때의 정확도와 channel impact factor를 고려한 BPSO를 적용시켰을 때의 정확도를 비교, 분석하였다.

• 대한전자공학회논문지SD
• /
• 제43권10호
• /
• pp.34-44
• /
• 2006

시스템 온 칩(SoC) 설계에서는 설계 시간을 줄이기 위해서 미리 검증된 IP(Intellectual Property) 하드웨어 블록을 사용한다. 서로 다른 통신 프로토콜을 사용하는 IP간의 연결을 위해서는 인터페이스 회로가 필요하다. SoC 설계에는 인터페이스 회로 설계가 빈번하게 발생하며, 오류발생 가능성이 높다. 따라서 인터페이스 회로의 자동생성이 필요하다. IP의 통신 프로토콜을 입력으로 인터페이스 회로를 생성하는 여러 연구가 진행되어 왔다. 그러나 IP의 통신 프로토콜만으로는 인터페이스 회로를 생성하기 어려운 경우가 있다. 주소를 사용하는 IP와 주소를 사용하지 않는 IP간 연결, 주소/데이터가 여러 포트로 전송되는 IP와 하나의 포트로 전송되는 IP간 연결, 매 전송마다 주소와 데이터가 함께 전송되는 IP와 하나의 시작주소에 여러 데이터가 전송되는 IP간의 연결 등에서는 기존 방법에 의한 인터페이스 회로의 자동 생성이 어렵다. 그것은 기존의 방법들이 IP 특성에 따른 범주 구분과, 특성간의 대응에 따라 합성 알고리즘도 변해야 함을 고려하지 않았기 때문이다. 본 논문에서는 IP의 통신프로토콜 특성에 따라 범주를 나누고, 어떤 특성을 갖는 IP의 연결인 지에 따라 인터페이스 합성 알고리즘이 이를 고려할 수 있도록 하였다. 실험에서는 다양한 특성을 갖는 IP 간 연결을 위한 인터페이스 회로를 생성하고 검증했음을 보인다.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 1999년도 추계학술대회-지능형 정보기술과 미래조직 Information Technology and Future Organization
• /
• pp.281-289
• /
• 1999

The objective of this paper is to propose a brand new interface mechanism to provide more intelligent decision making support for EC problems. Its main virtue is based on a numerical process mechanism by using an Extended AND-OR Graph (EAOG)-based logic algebra. Using this mechanism, decision makers engaged in electronic commerce (EC) can effectively deal with complicated decision making problems. In the field of traditional expert systems research, AND-OR Graph approach has been suggested as a useful tool for representing the logic flowchart of the forward and/or backward chaining inference methods. However, the AND-OR Graph approach cannot be effectively used in the EC problems in which real-time problem-solving property should be highly required. In this sense, we propose the EAOG inference mechanism for EC problem-solving in which heurisric knowledge necessary for intelligent EC problem-solving can be represented in a form of matrix. Finally, we have proved the validity of our approach with several propositions and an illustrative EC example

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2009년 추계학술대회논문집
• /
• pp.211-217
• /
• 2009

The numerical program has been developed for the purpose of the complicate geometries application using CIP method. The non-staggered, non-orthogonal, and unstructured grid system can be also used for the various geometries in the program. For validating CIP solver, the lid-driven cavity flow and solitary wave propagation flow are carried out. Test results show a good agreement with the verified results. The dynamic solver was used for the behavior of moving body. Interface process between the two solvers is introduced. The research was performed on the flow problem around torpedo and log and the flow problem in a tank in order to analyze the three phase flow problem Although the comparison to the verified results was not quantitatively performed, the trend of the results was reasonable.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2007년도 춘계학술대회 논문집
• /
• pp.1746-1751
• /
• 2007

The bullet train runs more than 200km/h so that it is invisible to the naked eye. For this reason, KTX solely dependent on the ATC operation system. Above all things, 'stability' and 'regular time' are essential to Train Control Operation system. But intermittently, there is a signal transmission problem between ATC operation facilities and ground equipments. which is caused by wheel sliding and friction coefficient between wheel and rail when the train starts. This paper analyzed train service data for one year for accurate survey. In addition, this thesis examines the main cause of a signal transmission problem through simulation and experiment. Further, the method to solve the conventional problem is presented.

• 산업공학
• /
• 제11권1호
• /
• pp.97-103
• /
• 1998

In these days many people have studied on the berthing problem. The operation rules differ from port so port and the problem is highly dependent on the environment of port. The purpose of this study is to develop a decision support system decision makers of the berthing problem for Inchon Port. The system is developed with graphic user interface(GUI) using user-interactive approach and some general and specific rules for Inchon Prot are considered. The system is composed of the following four parts ; the input/output part, the automatic berthing part by the system using rules, the manual berthing part by user, and the part for modifying results or handling exceptional events. The system is designed to assign ship to berths by matching the characteristics with environmental and operational constraints of Inchon Port. We expect that this system can provide decision makers with an efficient and fast way to berthing and can reduce wastes of time, space, and manpower in port operations.

• Journal of Mechanical Science and Technology
• /
• 제14권7호
• /
• pp.744-751
• /
• 2000

To achieve lower flying height for high areal recording density, the laser zone texturing of the disk needs to be designed to reduce glide height. One problem of the laser bump design is that the regular laser bump pattern often produces glide resonance phenomenon, which leads to failure of the glide height test. However, it was found in this study that glide resonance is an intrinsic problem of the glide head used and resonance phenomenon depends on the type of the head slider, that is, the natural frequency of the slider body. Therefore, higher glide height or glide failure caused by glide resonance does not lead to head/media interface problem in the real drive operating conditions in which the data head is used. Pseudo-random bump pattern greatly reduces the glide resonance. Smaller bump pitch will also help to reduce the glide resonance. However, as bump spacing becomes smaller, glide height will be increased due to increased air pressure developed around the bumps. Lowering bump height is the most effect way to reduce glide avalanche.

• 대한기계학회논문집B
• /
• 제24권6호
• /
• pp.792-801
• /
• 2000

The problem of phase change from liquid to solid in the inviscid plane-stagnation flow is theoretically investigated. The solution at the initial stage of freezing is obtained by expanding it in powers of time, and the final equilibrium state is determined from the steady-state governing equations. The transient solution is dependent on the three dimensionless parameters, but the equilibrium state is determined by one parameter of (temperature ratio/conductivity ratio). The effect of the fluid flow on the growth rate of the solid in the pure conduction problem can be clearly seen from the solution of the initial stage and the final equilibrium state. The characteristics of the transient heat transfer at the surface of the solid and the liquid side of the solid-liquid interface for all the dimensionless parameters are elucidated.