• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 2001년도 가을 학술발표회 논문집
• /
• pp.147-158
• /
• 2001

In general, two methods have been used to predict settlement of soft ground. One method is Terzaghi's one dimensional consolidation theory which gives time-settlement relationship using the standard consolidation test results. The other is forecasting method of ground settlement to be occured in the future using in-situ monitoring data. The above both methods have some defects in application manner or in itself especially in very deep and soft clayey ground. In view of the lessons and experiences of soft ground improvement projects, several techniques were proposed for more accurate theorectical calculation of consolidation settlement as follows ; ① Subdivision of soft ground, ② Consideration of secondary compression, ③ Using the modified compression index, etc. And also, revised hyperbolic fitting method was suggested to minimize the error of predicted future settlement. In addition, revised De-Beer equation of immediate settlement of loose sandy soil was proposed to overcome the tendency to show too small settlement calculation results by original De-Deer equation. And also, considering the various effects of settlement delay in the improved ground by vertical drains, time-settlement caculation equation(Onoue method) was revised to match the tendency of settlement delay by using the characteristics of discharge capacity decreases of vertical drain with time elapse by the pattern of hyperbolic equation.

• Steel and Composite Structures
• /
• 제2권2호
• /
• pp.147-160
• /
• 2002

This paper summarises results of the research performed at the Department of Steel Structures and Structural Mechanics from the "Politehnica" University of Timisoara, Romania, in order to evaluate the performance of beam-to-column extended end plate connections for steel and composite joints. It comprises laboratory tests on steel and composite joints, and numerical modelling of joints, based on tests. Tested joints are double-sided, with structural elements realised of welded steel sections. The columns are of cruciform cross-section, while the beams are of I section. Both monotonic and cyclic loading, symmetrically and antisymmetrically, has been applied. On the basis of tested joints, a refined computer model has been calibrated using a special connection element of the computer code DRAIN 2DX. In this way, a static/dynamic structural analysis of framed structures with real characteristics of the beam to column joints is possible.

• 한국전기전자재료학회논문지
• /
• 제20권6호
• /
• pp.487-490
• /
• 2007

In this paper, the formation and thermal stability characteristics of Ni silicide using Ni-V alloy on Boron cluster ($B_{18}H_{22}$) implanted bulk and SOI substrate were examined in comparison with pure Ni for nano-scale CMOSFET. The Ni silicide using Ni-V alloy on $B_{18}H_{22}$ implanted SOI substrate after high temperature post-silicidation annealing showed the lower sheet resistance, no agglomeration interface image and lower surface roughness than that using pure Ni. The thermal stability of Ni silicide was improved by using Ni-V alloy on $B_{18}H_{22}$ implanted SOI substrate.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2007년도 하계학술대회 논문집 Vol.8
• /
• pp.395-395
• /
• 2007

We report on the fabrication of soluble pentacene-based thin-film transistors (TFTs) that consist of $NiO_x$, poly-vinyl phenol (PVP), and Ni for the source-drain (SID) electrodes, gate dielectric, and gate electrode, respectively. The $NiO_x$ SID electrodes of which the work function is well matched to that of soluble pentacene are deposited on a soluble pentacenechannel by sputter deposited of NiO powder and show a moderately low but still effective transmittance of ~65% in the visible range along with a good sheet resistance of ${\sim}40{\Omega}/{\square}$. The maximum saturation current of our soluble pentacene-based TFT is about $15{\mu}A$ at a gate bias of -40showing a high field effect mobility of $0.06cm^2/Vs$ in the dark, and the on/off current ratio of our TFT is about $10^4$. It is concluded that jointly adopting $NiO_x$ for the S/D electrodes and PVP for gate dielectric realizes a high-quality soluble pentacene-based TFT.

• 전기전자학회논문지
• /
• 제25권1호
• /
• pp.15-24
• /
• 2021

본 논문에서는 CDT(Conventional Double Trench) MOSFET보다 스위칭 시간과 손실이 적은 1700 V EPDT(Extended P+ shielding floating gate Double Trench) MOSFET 구조를 제안하였다. 제안한 EPDT MOSFET 구조는 CDT MOSFET에서 소스 Trench의 P+ shielding 영역을 늘리고 게이트를 N+와 플로팅 P- 폴리실리콘 게이트로 나누었다. Sentaurus TCAD 시뮬레이션을 통해 두 구조를 비교한 결과 온 저항은 거의 차이가 없었으나 Crss(게이트-드레인 간 커패시턴스)는 게이트에 0 V 인가 시에는 CDT MOSFET 대비 32.54 % 줄었고 7 V 인가 시에는 65.5 % 감소하였다. 결과적으로 스위칭 시간 및 손실은 각각 45 %, 32.6 % 줄어 스위칭 특성이 크게 개선되었다.

• 반도체디스플레이기술학회지
• /
• 제6권3호
• /
• pp.19-23
• /
• 2007

The NiSi is very promising candidate for the metallization in 45 nm CMOS process such as FUSI(fully silicided) gate and source/drain contact because it exhibits non-size dependent resistance, low silicon consumption and mid-gap workfunction. Ni film was first deposited by using ALD (atomic layer deposition) technique with Bis-Ni precursor and $H_2$ reactant gas at $220^{\circ}C$ with deposition rate of $1.25\;{\AA}/cycle$. The as-deposited Ni film exhibited a sheet resistance of $5\;{\Omega}/{\square}$. RTP (repaid thermal process) was then performed by varying temperature from $400^{\circ}C$ to $900^{\circ}C$ in $N_2$ ambient for the formation of NiSi. The process temperature window for the formation of low-resistance NiSi was estimated from $600^{\circ}C$ to $800^{\circ}C$ and from $700^{\circ}C$ to $800^{\circ}C$ with and without Ti capping layer. The respective sheet resistance of the films was changed to $2.5\;{\Omega}/{\square}$ and $3\;{\Omega}/{\square}$ after silicidation. This is because Ti capping layer increases reaction between Ni and Si and suppresses the oxidation and impurity incorporation into Ni film during silicidation process. The NiSi films were treated by additional thermal stress in a resistively heated furnace for test of thermal stability, showing that the film heat-treated at $800^{\circ}C$ was more stable than that at $700^{\circ}C$ due to better crystallinity.

• Journal of Chest Surgery
• /
• 제33권6호
• /
• pp.445-468
• /
• 2000

Background: Retrograde cerebral perfusion(RCP) is currently used for brain protection during aorta surgery, however, for the safety of it, various data published so far are insufficient. We performed RCP using pig and investiaged various parameters of cerebral metabolism and brain injury after RCP under deep hypothermia. Material and Method: We used two experimental groups: in group I(7 pigs, 20 kg), we performed RCP for 120 minutes and in group II (5 pigs, 20 kg), we did it for 90 minutes. Nasopharyngeal temperature, jugular venous oxygen saturation, electroencephalogram were continuously monitored, and we checked the parameters of cerebral metabolism, histological changes and serum levels of neuron-specific enolose(NSE) and lactic dehydrogenase(LDH). Central venous pressure during RCP was mainained in the range of 25 to 30 mmHg. Result: Perfusion flow rates(ml/min) during RCP were 130$\pm$57.7(30 minutes), 108.6$\pm$55.2(60 minutes), 107.1$\pm$58.8(90 minutes), 98.6$\pm$58.7(120 minutes) in group I and 72$\pm$11.0(30 minutes), 72$\pm$11.0(60 minutes), 74$\pm$11.4(90 minutes) in group II. The ratios of drain flow to perfusion flow were 0.18(30 minutes), 0.19(60 minutes), 0.17(90 minutes), 0.16(120 minutes) in group I and 0.21, 0.20, 0.17 in group II. Oxygen consumptions(ml/min) during RCP were 1.80$\pm$1.37(30 minutes), 1.72$\pm$1.23(60 minutes), 1.38$\pm$0.82(90 minutes), 1.18$\pm$0.67(120 minutes) in group I and 1.56$\pm$0.28(30 minutes), 1.25$\pm$0.28(60 minutes), 1.13$\pm$0.26(90 minutes). We could observe an decreasing tendency of oxygen consumption after 90 minutes of RCP in group I. Cerebrovascular resistance(dynes.sec.cm-5) during RCP in group I incrased from 71370.9$\pm$369145.5 to 83920.9$\pm$49949.0 after the time frame of 90 minutes(p<0.05). Lactate(mg/min) appeared after 30 minutes of RCP and the levels were 0.15$\pm$0.07(30 minutes), 0.18$\pm$0.10(60 minutes), 0.19$\pm$0.19(90 minutes), 0.18$\pm$0.10(120 minutes) in group I and 0.13$\pm$0.09(30 minutes), 0.19$\pm$0.03(60 minutes), 0.29$\pm$0.11(90 minutes) in group II. Glucose utilization, exudation of carbon dioxide, differences of cerebral tissue acidosis between perfusion blood and drain blood were maintained constantly during RCP. Oxygen saturation levels(%) in drain blood during RCP were 22.9$\pm$4.4(30 minutes), 19.2$\pm$4.5(60 minutes), 17.7$\pm$2.8(90 minutes), 14.9$\pm$2.8(120 minutes) in group I and 21.3$\pm$8.6(30 minutes), 20.8$\pm$17.6(60 minutes), 21.1$\pm$12.1(90 minutes) in group II. There were no significant changes in cerebral metabolic parameters between two groups. Differences in serum levels of NSE and LDH between perfusion blood and drain blood during RCP showed no statistical significance. Serum levels of NSE and LDH after resuming of cardipulmonary bypass decreased to the level before RCP. Brain water contents were 0.73$\pm$0.03 in group I and 0.69$\pm$0.06 in group II and were higher than those of the controls(p<0.05). The light microscopic findings of cerebral neocortex, basal ganglia, hippocampus(CA1 region) and cerebellum showed no evidence of cerebral injury in two groups and there were no different electron microscopy in both groups(neocortex, basal ganglia and hippocampus), but they were thought to be reversible findings. Conclusion: Although we did not proceed this study after survival of pigs, we could perform the RCP successfully for 120 minutes with minimal cerebral metabolism and no evidence of irreversible brain damage. The results of NSE and LDH during and after RCP should be reevaluated with survival data.

• 한국전자파학회논문지
• /
• 제24권11호
• /
• pp.1064-1073
• /
• 2013

본 논문에서는 내부 고조파 조정 회로로 구성되는 셀룰러와 L-대역용 소형의 고효율 370 W GaN(Gallium Nitride) HEMT(High Electron Mobility Transistor) 소형 전력 증폭기(PA)를 구현하였다. 원천 및 2차 고조파 주파수에서 동시에 높은 효율을 내기 위해 새로운 회로 정합 형태를 적용했다. 소형화를 위하여 새로운 41.8 mm GaN HEMT와 2개의 MOS(Metal Oxide Semiconductor) 캐패시터를 구성 물질의 변화를 이용하여 열 저항을 개선한 $10.16{\times}10.16{\times}1.5Tmm^3$ 크기의 새로운 패키지에 와이어 본딩으로 결합하였다. 드레인 바이어스 48 V 인가 시, 개발된 GaN HEMT 전력 증폭기는 370 W 포화 출력 전력(Psat.)과 770~870 MHz에서 80 % 이상, 1,805~1,880 MHz에서 75 % 이상의 드레인 효율(DE)을 나타내었다. 이는 지금까지 보고된 셀룰러와 L대역에서 GaN HEMT 전력 증폭기 중 최고의 효율과 출력 전력 특성이다.

• 대한전자공학회논문지SD
• /
• 제38권2호
• /
• pp.95-102
• /
• 2001

Nano-scale의 게이트 길이를 가지는 MOSFET소자는 접합 깊이가 20∼30㎚정도로 매우 얕은 소스/드레인 확장 영역을 필요로 한다. 본 연구에서는 As₂/sup +/ 이온의 10keV이하의 낮은 에너지 이온 주입과 RTA(rapid thermal annealing)공정을 적용하여 20㎚이하의 얕은 접합 깊이와 1.O㏀/□ 이하의 낮은 면저항 값을 가지는 n/sup +/-p접합을 구현 하였다. 이렇게 형성된 n/sup +/-p 접합을 nano-scale MOSFET소자 제작에 적용 시켜서 70㎚의 게이트 길이를 가지는 NMOSFET을 제작하였다. 소스/드레인 확장 영역을 As₂/sup +/ 5keV의 이온 주입으로 형성한 100㎚의 게이트 길이를 가지는 NMOSFET의 경우, 60mV의 낮은 V/sub T/(문턱 전압감소) 와 87.2㎷의 DIBL (drain induced barrier lowering) 특성을 확인하였다. 10/sup 20/㎝/sup -3/이상의 도핑 농도를 가진 abrupt한 20㎚급의 얕은 접합, 그리고 이러한 접합이 적용된 NMOSFET소자의 전기적 특성들은 As₂/sup +/의 낮은 에너지의 이온 주입 기술이 nano-scale NMOSFET소자 제작에 적용될 수 있다는 것을 제시한다.

• 대한전자공학회논문지SD
• /
• 제41권12호
• /
• pp.27-34
• /
• 2004

최근 실리콘 미세공정의 발달로 상용화된 0.2$\mum$ 게이트길이 이하의 deep submicron MOSFET 출력특성을 정확히 모델링하기 위해서는 RF 기판 회로 연구가 필수적이다. 먼저 본 논문에서는 기판 캐패시던스와 기판 저항이 병렬로 연결된 모델과 기판 저항만을 사용한 단순 모델들에 적합한 직접 추출 방법을 각각 개발하였다. 이 추출방법들을 0.15$\mum$ CMOS 소자에 적용한 결과 단순 모델보다 RC 병렬 기판모델이 측정된 $Y_{22}$-parameter에 30GHz까지 더 잘 일치하는 것을 확인하였으며, 이는 RC 병렬 기판모델 및 직접추출방법의 RF 정확도를 증명한다. 이러한 RC 병렬 기판모델을 사용하여 게이트 길이를 0.11에서 0.5$\mum$까지 변화시키고 드레인 전압을 0에서 1.2V까지 증가시키면서 기판 모델 파라미터들의 bias 종속 특성과 게이트 길이 종속 특성을 새롭게 추출하였다. 이러한 새로운 추출 결과는 scalable한 RF 비선형 기판 모델 개발에 유용하게 사용될 것이다.