• 한국신재생에너지학회:학술대회논문집
• /
• 한국신재생에너지학회 2008년도 춘계학술대회 논문집
• /
• pp.507-510
• /
• 2008

Characteristics of an autothermal reformer at various operating parameters have been studied in this paper. Numerical method has been used, and simulation model has been developed for the analysis. Full Combustion reaction, Steam Reforming(SR) reaction, Water-Gas Shift(WGS) reaction, and Direct Steam Reforming(DSR) reaction are assumed as dominant chemical reactions in the autothermal reformer. Simulation results are compared with experimental results for code validation. Operating parameters of the autothermal reformer are inlet temperature, Oxygen to Carbon Ratio(OCR), Steam to Carbon Ratio(SCR), and Gas Hourly Space Veolcity(GHSV). SR reaction rate decreases with low inlet temperature. If OCR is increased, $H_2$ yield is increased but optimal point is suggested. WGS reaction is activated with high SCR. When GHSV is increased, reforming efficiency is increased but pressure drop may decrease the system efficiency.

• 대한기계학회논문집B
• /
• 제28권12호
• /
• pp.1501-1510
• /
• 2004

The pumping characteristics of a single-stage disk-type drag pump (DTDP) are calculated for the variation of the vertical clearance between a rotor and stator by the three-dimensional direct simulation Monte Carlo (DSMC) method. The gas flow mainly belongs to the molecular transition flow region. Spiral channels of a DTDP are cut on the both the upper and lower sides of a rotating disk, but a stationary disk is planar. The interaction between molecules is described by the variable hard-sphere model. The no time counter method is used as a collision sampling technique. The vertical clearance has a significant effect on the pumping performance. Experiments are performed under the outlet pressure range of 0.4∼533 Pa. When the numerical results are compared with the experimental data, the numerical results agree well quantitatively

• 한국광학회:학술대회논문집
• /
• 한국광학회 2003년도 제14회 정기총회 및 03년 동계학술발표회
• /
• pp.184-185
• /
• 2003

Unlike the conventional waveguide such as optical fiber using total internal reflection, photonic crystal waveguide(PCW), a waveguide made of a line defect in a photonic crystal(PC) structure, does not admit an analytic approach due to its complexity but requires a direct numerical approach. Here, we present numerical results of computer simulation for PCW by using the three-dimensional(3D) Finite-Difference Time -Domain(FDTD) algorithm. (omitted)

• 한국수자원학회논문집
• /
• 제56권spc1호
• /
• pp.1059-1069
• /
• 2023

댐붕괴흐름은 댐이 갑자기 붕괴하여 제어가 어려운 상태의 고속흐름이 방출되는 현상이다. 이 연구에서는 3차원의 댐붕괴흐름을 모의하기 위해 OpenFOAM을 사용하여 층류 및 난류 모델을 적용하고 그 결과를 비교하였다. 난류 모의를 위해 레이놀즈 평균 나비에-스토크스 (Reynolds-Averaged Navier-Stokes) 모델, 구체적으로 k-ε 모델을 사용하였다. 수리모형실험과 함께 수정된 다층 블록 장애물 시나리오를 대상으로 두 가지 모델을 평가하였다. 두 모델 모두 댐붕괴흐름을 효과적으로 재현하였으며, 난류 모델은 흐름의 변동성을 감소시키는 역할을 보여줬다. 그러나 난류 모델에서의 과도한 에너지소산은 수위를 과소 평가하게 하는 것으로 나타났다. 수치기법 및 격자 해상도를 개선하여 적용한 결과 흐름재현성이 향상되었는데 이는 특히 구조물 근처의 난류흐름에서 두드러졌다. 모델 안정성의 경우 난류모델의 사용여부보다는 수치기법 및 격자 해상도의 개선에 더 크게 영향을 받았다. k-ε 모델에 내재된 시간평균처리의 특성은 불연속성과 불안정성이 두드러진 댐붕괴흐름을 재현하는 데 한계가 있음을 나타냈다. RANS 모델을 포함한 난류모의는 방대한 계산자원이 필요하지만, 층류 모델과 비교하여 성능 향상이 제한적이었다. 댐붕괴흐름을 정확히 재현하기 위해 LES (Large Eddy Simulation) 및 DNS (Direct Numerical Simulation)과 같은 고급 난류 모델의 사용이 권장되며, 이를 위해서는 미세한 공간 및 시간 스케일의 구성이 필수적이다. 이 연구를 통해 댐붕괴흐름을 모의할 때 기본적으로 사용할 수 있는 주요 접근법과 적용가능성을 측정하였으며, 구조물 근처에서 난류흐름에 대한 정확한 표현의 중요성을 강조할 수 있었다.

• 비파괴검사학회지
• /
• 제28권6호
• /
• pp.483-495
• /
• 2008

The paper presents a numerical study to evaluate the phase velocities and attenuations of the average longitudinal and shear ultrasonic waves resulting from multiple scattering in fiber-reinforced composites. A computational procedure developed in this work is first used to produce a random, yet largely even distribution of fibers. Both the viscoelastic epoxy matrix and lossless randomly distributed graphite fibers are modeled using the mass-spring-dashpot lattice model, with no damping for the latter. By numerically simulating ultrasonic through-transmission tests using this direct model of composites, phase velocities and attenuations of the longitudinal and shear waves through the composite are found as functions of frequency or fiber concentration. The numerical results are observed to generally agree with the corresponding results in the literature. Discrepancies found in some detail aspects, particularly in the attenuation results, are also addressed.

• 한국가시화정보학회지
• /
• 제15권3호
• /
• pp.27-33
• /
• 2017

Direct numerical simulations of turbulent channel flows with moving wall conditions on the top wall are performed to examine the effects of the moving wall on the turbulent characteristics. The moving wall velocity only applied to the top wall with the opposite direction to the main flow is systematically varied to reveal the sustained-mechanism for turbulence. The turbulence statistics for the Couette-Poiseuille flow, such as mean velocity, root mean square of the velocity fluctuations, Reynolds shear stress and pre-multiplied energy spectra of the velocity fluctuations, are compared with those of canonical turbulent channel flows. The comparison suggests that although the turbulent activity on the top wall increases with increasing the Reynolds number, that on the bottom wall decreases, contrary to the previous finding for the canonical turbulent channel flows. The increase of the turbulent energy on the top wall is attributed to not only the increase of the Reynolds number but also elongation of the logarithmic layer due to increase of the wall layer on the top wall. However, because the logarithmic layer is shortened on the bottom wall due to the decrease of the wall layer, the turbulence energy on the bottom wall decreases despite of the increase of the Reynolds number.

• 한국추진공학회:학술대회논문집
• /
• 한국추진공학회 2008년 영문 학술대회
• /
• pp.45-52
• /
• 2008

In order to investigate the initiation and propagation processes of a spherical detonation wave induced by direct initiation, numerical simulations were carried out using two-dimensional compressible Euler equations with an axisymmetric assumption and a one-step reaction model based on Arrhenius kinetics with various levels of ignition energy. By varying the amount of ignition energy, three typical initiation behaviors, which were subcritical, supercritical and critical regimes, were observed. Then, the ignition energy of more than $137.5{\times}10^6$ in non-dimensional value was required for initiating a spherical detonation wave, and the minimum ignition energy(i.e., critical energy) was less than that of the one-dimensional simulation reported by a previous numerical work. When the ignition energy was less than the critical energy, the blast wave generated from an ignition source continued to attenuate due to the separation of the blast wave and a reaction front. Therefore, detonation was not initiated in the subcrtical regime. When the ignition energy was more than the minimum initiation energy, the blast wave developed into a multiheaded detonation wave propagating spherically at CJ velocity, and then a cellular pattern radiated regularly out from the ignition center in the supercritical regime. The influence on ignition energy was observed in the cell width near the ignition center, but the cell width on the fully developed detonation remained constant during the expanding of detonation wave due to the consecutive formation of new triple points, regardless of ignition energy. When the ignition energy was equal to the critical energy, the decoupling of the blast wave and a reaction front appeared, as occurred in the subcrtical regime. After that, the detonation bubble induced by the local explosion behind the blast wave expanded and developed into the multiheaded detonation wave in the critical regime. Although few triple points were observed in the vicinity of the ignition core, the regularly located cellular pattern was generated after the onset of the multiheaded detonation. Then, the average cell width on the fully developed detonation was almost to that in the supercritical regime. These numerical results qualitatively agreed with previous experimental works regarding the initiation and propagation processes.

• 한국전기전자재료학회논문지
• /
• 제35권6호
• /
• pp.616-621
• /
• 2022

Numerical simulation is a good way to predict the conversion efficiency of solar cells without a direct experimentation and to achieve low cost and high efficiency through optimizing each step of solar cell fabrication. TOPCon industrial solar cells fabricated with n-type silicon wafers on a larger area have achieved a higher efficiency than p-type TOPCon solar cells. Electrical and optical losses of the front surface are the main factors limiting the efficiency of the solar cell. In this work, an optimization of boron-doped emitter surface and front electrodes through numerical simulation using "Griddler" is reported. Through the analysis of the results of simulation, it was confirmed that the emitter sheet resistance of 150 Ω/sq along the front electrodes having a finger width of 20 ㎛, and the number of finger lines ~130 for silicon wafer of M6 size is an optimized technology for the front emitter surface of the n-type TOPCon solar cells that can be developed.

• Structural Engineering and Mechanics
• /
• 제32권1호
• /
• pp.1-20
• /
• 2009

The present contribution addresses uncertainty quantification and uncertainty propagation in structural mechanics using stochastic analysis. Presently available procedures to describe uncertainties in load and resistance within a suitable mathematical framework are shortly addressed. Monte Carlo methods are proposed for studying the variability in the structural properties and for their propagation to the response. The general applicability and versatility of Monte Carlo Simulation is demonstrated in the context with computational models that have been developed for deterministic structural analysis. After discussing Direct Monte Carlo Simulation for the assessment of the response variability, some recently developed advanced Monte Carlo methods applied for reliability assessment are described, such as Importance Sampling for linear uncertain structures subjected to Gaussian loading, Line Sampling in linear dynamics and Subset simulation. The numerical example demonstrates the applicability of Line Sampling to general linear uncertain FE systems under Gaussian distributed excitation.

• Structural Engineering and Mechanics
• /
• 제66권3호
• /
• pp.379-386
• /
• 2018

The present study addresses the direct and indirect methods of determining the mode-I fracture toughness of concrete using experimental tests and particle flow code. The direct method used is compaction tensile test and the indirect methods are notched Brazilian disc test, semi-circular bend specimen test, and hollow center cracked disc. The experiments were carried out to determine which indirect method yields the fracture toughness closer to the one obtained by the direct method. In the numerical analysis, the PFC model was first calibrated with respect to the data obtained from the Brazilian laboratory test. The crack paths observed in the simulated tests were in reasonable accordance with experimental results. The discrete element simulations demonstrated that the macro fractures in the models are caused by microscopic tensile breakages on large numbers of bonded particles. The mode-I fracture toughness in the direct tensile test was smaller than the indirect testing results. The fracture toughness obtained from the SCB test was closer to the direct test results. Hence, the semi-circular bend test is recommended as a proper experiment for determination of mode-I fracture toughness of concrete in the absence of direct tests.