• 대한기계학회논문집B
• /
• 제41권2호
• /
• pp.117-124
• /
• 2017

본 연구에서는 Coloring 기법을 적용한 Gauss-Seidel 해법의 연산 성능을 분석하기 위해 2차원과 3차원 전도 열전달 문제를 다양한 격자 크기에서 해석하였다. 지배방정식의 이산화는 유한차분법과 유한요소법을 사용하였다. CPU의 경우에는 상대적으로 작은 격자계에서 연산 성능이 좋으며, 계산에 사용되는 메모리의 크기가 캐시메모리보다 크게 되면 연산 성능이 급격히 떨어진다. 반면에, GPU는 메모리 지연시간 숨김 특성으로 인하여 격자의 수가 충분히 많을 때 연산 성능이 좋다. GPU에 기반한 Colored Gauss-Seidel 해법은 단일 CPU를 이용한 연산에 비해서 각각 최대 7배의 속도 향상을 보인다. 또한, GPU 기반에서 Colored Gauss-Seidel 해법은 Jacobi 보다 약 2배 빠름을 확인하였다.

• 대한치과기공학회지
• /
• 제40권4호
• /
• pp.217-224
• /
• 2018

Purpose: The aim of this study was to evaluate accuracy of three type complete mandibular denture of before and after polymerization. Methods: Mandibular edentulous model was selected as the master model. 15 study models were made by Type IV stone. Wax complete mandibular dentures were produced by the denture base and artificial teeth. Before and after curing, STL files were obtained using a blue scanner. By superimposing the digitized complete mandibular denture data(after curing) with the CAD-reference(before curing) three-dimensionally, visual fit-discrepancies were drawn by calculating the root mean square (RMS) and visualized on a color-difference map. Each calculated RMS-value was statistically analyzed by 1-way analysis of variance(ANOVA) (${\alpha}=.05$). Results: Mean(SD) RMS-values was OM group $88.98(6.10){\mu}m$, BM group $82.35(13.46){\mu}m$, BDM group $77.83(9.46){\mu}m$. The results of the 1-way ANOVA showed no statistically significant differences in the RMS values of the Three groups for the material (P > .241). Conclusion : Deformation of artificial teeth position was observed in all groups after resin polymerization. But the values, all group were within the clinically acceptable range. The values of BDM group showed the least deformation than the other two groups.

• 한국수자원학회논문집
• /
• 제50권6호
• /
• pp.387-395
• /
• 2017

기후변화에 따른 해수면 상승으로 인하여 해안지역의 지하수계에 해수침투가 가중된다. 지하수의 염분농도가 증가하면 지하수면 상부의 불포화 토양에서도 염분 농도가 증가할 수 있으며, 이는 농경지에서 작물피해를 일으킬 수 있다. 해수면이 상승함에 따라 내륙의 지하수위도 함께 상승한다. 이는 불포화 토양층의 두께를 감소시켜 해안 저지대의 경작에 피해를 끼칠 수 있다. 본 연구에서 지하수 해수침투는 3차원 모델, 토양 염류화 평가는 연직 1차원 모델을 합성 적용하여 해안 농경지에 대한 해수면 상승 피해를 평가하는 방법을 개발하였다. 3차원 해수침투 모델에서 지하수면의 수위와 농도분포를 계산하고 최상부 절점 중에서 염분 농도가 기준 값 이상인 절점에서 지하수면과 지표면 사이의 토양층에서 연직 1차원 모델링으로 토양층의 염분 농도와 불포화대 두께를 계산하였다. 농경지의 토양 염류화는 작물의 뿌리 심도에서 보통 작물의 생육한계 염분농도를 기준으로 판단하였다. 개발된 모델링 방법을 가상의 간척농경지에 적용하였다. 해수면 상승자료로 IPCC의 RCP 4.5와 8.5 시나리오를 사용하였다. 평가 결과는 2050년과 2100년에 대하여 제시하였다. 연구결과 대상지역에서 기후변화 시나리오 RCP 8.5에서 2100년에는 지하수 염류화 피해 면적은 간척지 육지면적 대비 7.8%, 염류화 토양 면적은 6.0%, 불포화층의 두께가 뿌리심도보다 적은 지역의 면적은 1.6% 증가하는 것으로 분석되었다.

• Transactions on Electrical and Electronic Materials
• /
• 제13권5호
• /
• pp.256-261
• /
• 2012

Recently, many photovoltaic systems (PV systems) including solar parks and PV farms have been built to prepare for the post fossil fuel era. To investigate the degradation process of the PV systems and thus, efficiently operate PV systems, there is a need to visually monitor PV systems in the range of infrared ray through the Internet. For efficient visual monitoring, this paper explores a multithreaded implementation of a recently developed HEVC standard whose compression efficiency is almost two times higher than H.264. For an efficient parallel implementation under a meshbased 64 multicore system, this work takes into account various design choices which can solve potential problems of a two-dimensional interconnects-based 64 multicore system. These problems may have not occurred in a small-scale multicore system based on a simple bus network. Through extensive evaluation, this paper shows that, for an efficient multithreaded implementation of HEVC intra prediction in a mesh-based multicore system, much effort needs to be made to optimize communications among processing cores. Thus, this work provides three design choices regarding communications, i.e., main thread core location, cache home policy, and maximum coding unit size. These design choices are shown to improve the overall parallel performance of the HEVC intra prediction algorithm by up to 42%, achieving a 7 times higher speed-up.

• International Journal of CAD/CAM
• /
• 제5권1호
• /
• pp.19-27
• /
• 2005

As three-dimensional range scanners make large point clouds a more common initial representation of real world objects, a need arises for algorithms that can efficiently process point sets. In this paper, we present a method for extracting smooth surfaces from dense point clouds. Given an unorganized set of points in space as input, our algorithm first uses principal component analysis to estimate the surface variation at each point. After defining conditions for determining the geometric compatibility of a point and a surface, we examine the points in order of increasing surface variation to find points whose neighborhoods can be closely approximated by a single surface. These neighborhoods become seed regions for region growing. The region growing step clusters points that are geometrically compatible with the approximating surface and refines the surface as the region grows to obtain the best approximation of the largest number of points. When no more points can be added to a region, the algorithm stores the extracted surface. Our algorithm works quickly with little user interaction and requires a fraction of the memory needed for a standard mesh data structure. To demonstrate its usefulness, we show results on large point clouds acquired from real-world objects.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제20권3호
• /
• pp.243-259
• /
• 2016

The Richards equation for water movement in unsaturated soil is highly nonlinear partial differential equations which are not solvable analytically unless unrealistic and oversimplifying assumptions are made regarding the attributes, dynamics, and properties of the physical systems. Therefore, conventionally, numerical solutions are the only feasible procedures to model flow in partially saturated porous media. The standard Finite element numerical technique is usually coupled with an Euler time discretizations scheme. Except for the fully explicit forward method, any other Euler time-marching algorithm generates nonlinear algebraic equations which should be solved using iterative procedures such as Newton and Picard iterations. In this study, lumped mass and distributed mass in the frame of Picard and Newton iterative techniques were evaluated to determine the most efficient method to solve the Richards equation with finite element model. The accuracy and computational efficiency of the scheme and of the Picard and Newton models are assessed for three test problems simulating one-dimensional flow processes in unsaturated porous media. Results demonstrated that, the conventional mass distributed finite element method suffers from numerical oscillations at the wetting front, especially for very dry initial conditions. Even though small mesh sizes are applied for all the test problems, it is shown that the traditional mass-distributed scheme can still generate an incorrect response due to the highly nonlinear properties of water flow in unsaturated soil and cause numerical oscillation. On the other hand, non oscillatory solutions are obtained and non-physics solutions for these problems are evaded by using the mass-lumped finite element method.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제10권5호
• /
• pp.566-582
• /
• 2018

A moored barge alongside quay can be influenced by a nearby passing ship and its ship-generated waves. In this study, a time-domain numerical method based on a three-dimensional potential flow solver is developed to investigate the passing ship problem with a moored barge alongside quay. Potential flows around the passing ship and the moored barge alongside a quay is directly solved by using a classical finite element method. Total computational meshes including a passing ship, a moored barge and a quay is updated at each step with an efficient re-mesh algorithm. To validate the developed numerical method, a conventional ship wave problem and a passing ship problem on the open sea has been solved and the solutions are compared with the existing data. Then, a series of numerical computations were carried out to investigate the passing ship effect on a moored barge alongside quay. The characteristics of the passing ship effects are studied with varying the simulation parameters such as passing ship speed, separation distance, wall distances and waves. Focus is made on hydrodynamic forces due to the passing ship effect and its ship waves.

• 대한기계학회논문집A
• /
• 제24권1호
• /
• pp.144-155
• /
• 2000

초록 The validity, of a single parameter such as stress intensity, factor K or J-integral in traditional fracture mechanics depends strongly on the geometry, and loading condition. Therefore the second parameter like T-stress measuring the stress constraint is additionally needed to characterize the general crack-tip fields. While many, research works have been done to verify, the J-T description of elastic-plastic crack-tip stress fields in plane strain specimens, limited works (especially. for bimaterials) have been performed to describe the structural surface crack-front stress fields with the two parameters. On this background, via detailed three dimensional finite element analyses for surface-cracked plates and straight pipes of homogeneous materials and bimaterials under various loadings, we investigate the extended validity or limitation of the two parameter approach. We here first develop a full 3D mesh generating program for semi-elliptical surface cracks, and calculate elastic T-stress from the obtained finite element stress field. Comparing the J-T predictions to the elastic-plastic stresses from 3D finite element analyses. we then confirm the extended validity of fracture mechanics methodology based on the J-T two parameters in characterizing the surface crack-front fields of welded plates and pipes under various loadings.

• 한국자동차공학회논문집
• /
• 제14권4호
• /
• pp.39-48
• /
• 2006

In this research, in order to study the spray, combustion, and emission characteristics of the common rail DME engine, the target engine was disassembled, and 3D CAD file was constructed using a 3D measurement machine and a rapid prototyping machine. Using the obtained 3D geometry, fine moving meshes are generated, and three dimensional non-steady turbulence flow field and combustion phenomenon including spray were numerically analyzed. As a result, IMEP of DME and diesel in medium and high speed revolution showed similar performance. As the DME fuel start to burn in spray area, the vaporized fuel rapidly spreads squish area in low speed revolution. In the case of DME engine, CO and NOx are relatively consistent with experiment results. It was found that the break-up, evaporation, collision model of DME fuel need to be properly adjusted through matching the characteristics of fuel and injector for further improvement.

• International Journal of Aeronautical and Space Sciences
• /
• 제12권2호
• /
• pp.190-199
• /
• 2011

Unsteady flow simulations of complete helicopter configurations were conducted, and the flow fields and the aerodynamic interferences between the main rotor, fuselage, and tail rotor were investigated. For these simulations, a three-dimensional flow solver based on unstructured meshes was used, coupled with an overset mesh technique to handle relative motion among those components. To validate the flow solver, calculations were made for a UH-60A complete helicopter configuration at high-speed and low-speed forward flight conditions, and the unsteady airloads on the main rotor blade were compared to available flight test data and other calculated results. The results showed that the fuselage changed the rotor inflow distribution in the main rotor blade airloads. Such unsteady vibratory airloads were produced on the fuselage, which were nearly in-phase with the blade passage over the fuselage. The flow solver was then applied to the simulation of a generic complete helicopter configuration at various flight conditions, and the results were compared with those of the CAMRAD-II comprehensive analysis code. It was found that the main rotor blades strongly interact with a pair of disk-vortices at the outer edge of the rotor disk plane, which leads to high pulse airloads on the blade, and these airloads behave differently depending on the specific flight condition.