• Journal of Magnetics
• /
• 제18권3호
• /
• pp.342-347
• /
• 2013

Local wall thinning in pipelines affects the structural integrity of industries, such as nuclear power plants (NPPs). In the present study, a development of pulsed eddy current (PEC) technology that detects the wall thinning of pipelines covered with insulation is reviewed. The methods and experimental results, which have two kinds of probe with a single and double core, were compared. For this purpose, the single and double core probes having one and two excitation coils have been devised, and the differential probe with two Hall sensors has been fabricated to measure the wall thinning in insulated pipelines. The test sample is a stainless steel having different thickness, laminated by plastic insulation to simulate the pipelines in NPPs. The excitation coils in the probe is driven by a rectangular current pulse, the difference of two Hall sensors has been measured as a resultant PEC signal. The peak value of the detected signal is used to describe the wall thinning. The double core probe has better performance to detect the wall thinning covered with insulation; the single core probe can detect the wall thinning up to an insulation thickness of 18 mm, whereas the double probe can detect up to 25 mm. The results show that the double core PEC probe has the potential to detect the wall thinning in an insulated pipeline of the NPPs.

• Structural Engineering and Mechanics
• /
• 제83권6호
• /
• pp.709-727
• /
• 2022

Fastening systems have a significant role in the response of railway slab track systems. Although experimental tests indicate nonlinear behavior of fastening systems, they have been simulated as a linear spring-dashpot element in the available literature. In this paper, the influence of the nonlinear behavior of fastening systems on the slab track response was investigated. In this regard, a nonlinear model of vehicle/slab track interaction, including two commonly used fastening systems (i.e., RFFS and RWFS), was developed. The time history of excitation frequency of the fastening system was derived using the short time Fourier transform. The model was validated, using the results of a comprehensive field test carried out in this study. The frequency response of the track was studied to evaluate the effect of excitation frequency on the railway track response. The results obtained from the model were compared with those of the conventional linear model of vehicle/slab track interaction. The effects of vehicle speed, axle load, pad stiffness, fastening preload on the difference between the outputs obtained from the linear and nonlinear models were investigated through a parametric study. It was shown that the difference between the results obtained from linear and nonlinear models is up to 38 and 18 percent for RWFS and RFFS, respectively. Based on the outcomes obtained, a nonlinear to linear correction factor as a function of vehicle speed, vehicle axle load, pad stiffness and preload was derived. It was shown that consideration of the correction factor compensates the errors caused by the assumption of linear behavior for the fastening systems in the currently used vehicle track interaction models.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권5호
• /
• pp.1431-1445
• /
• 2022

We construct a deep neural network model named ECGResNet. This model can diagnosis diseases based on 12-lead ECG data of eight common cardiovascular diseases with a high accuracy. We chose the 16 Blocks of ResNet50 as the main body of the model and added the Squeeze-and-Excitation module to learn the data information between channels adaptively. We modified the first convolutional layer of ResNet50 which has a convolutional kernel of 7 to a superposition of convolutional kernels of 8 and 16 as our feature extraction method. This way allows the model to focus on the overall trend of the ECG signal while also noticing subtle changes. The model further improves the accuracy of cardiovascular and cerebrovascular disease classification by using a fully connected layer that integrates factors such as gender and age. The ECGResNet model adds Dropout layers to both the residual block and SE module of ResNet50, further avoiding the phenomenon of model overfitting. The model was eventually trained using a five-fold cross-validation and Flooding training method, with an accuracy of 95% on the test set and an F1-score of 0.841.We design a new deep neural network, innovate a multi-scale feature extraction method, and apply the SE module to extract features of ECG data.

• 한국건축시공학회지
• /
• 제23권1호
• /
• pp.11-22
• /
• 2023

본 연구에서는 디스크형 공시체를 활용하여 콘크리트의 깊이별 열화도를 평가하기 위한 방법을 개발하고자 하였다. 이를 위해 원통형 콘크리트 시편의 동탄성계수는 양단자유공진주기법을, 디스크형 콘크리트 시편의 동탄성계수는 임펄스 기법과 충격공진기법을 활용하여 측정하였고, 이를 비교 분석하였다. 실험결과, 임펄스 기법 및 충격공진기법 모두 동일한 지름일 경우 두께가 두꺼워지면 디스크 시편의 동탄성계수는 원주형 공시체의 동탄성계수에 가깝게 측정되었으며, 그 값의 변동성 은 줄어들었다. 또한 같은 (두께)/(반지름) 비율일 경우 지름이 증가하면 디스크 시편의 동탄성계수 측정값의 변동성은 감소하였고, 이러한 경향은 충격공진기법을 이용한 측정에서 더욱 분명하게 나타났다. 디스크의 동탄성계수 측정 시 지름 100mm, 두께25mm의 시편에 대해 충격공진기법을 이용하는 것이 오차율을 줄일 수 있는 방법으로 확인되었다.

• 한국지진공학회논문집
• /
• 제14권3호
• /
• pp.23-30
• /
• 2010

원자력발전소에 설치되는 안전관련 캐비닛형 전기기기는 설치 전에 내진검증이 요구된다. 전기기기의 동특성분석은 내진 검증에 포함된 중요한 과정이며, 기기의 정확한 해석모델을 작성하기 위해서도 필수적으로 요구되는 업무이다. 이 연구에서는 입력진동수준에 따른 기기의 동특성 변화를 분석하기 위하여 원전 지진감시시스템 캐비닛을 대상으로 진동대시험을 수행하고, 입력진동운동의 수준별로 계측된 진동응답신호를 진동수영역분해법으로 분석하였다. 분석결과, 대상기기는 입력진동수준의 크기에 따라 동특성이 비선형적으로 변화하고, 국내 원전의 안전정지지진 수준 이하의 진동에서도 동특성이 비선형적 거동을 보이고 있음을 확인하였다. 이러한 입력진동 수준에 따라 전기기기의 동특성이 비선형적으로 변하는 원인은 대상기기의 특성과 입력진동수준을 고려할 때 일반적인 재료 비선형보다는 각 부품들의 마찰력과 기하학적인 비선형성에 기인하는 것으로 판단된다. 따라서 전기 캐비닛의 입력진동수준에 따른 동특성의 비선형적 변화는 향후 안전관련 기기의 내진검증 업무에서 중요하게 검토되어야 할 것으로 판단된다.

• 대한토목학회논문집
• /
• 제30권6A호
• /
• pp.535-542
• /
• 2010

철도교량과 같이 주기적 가진에 의해 공진발생의 가능성이 높은 구조물에 있어서, 정확한 구조물의 동적특성치 파악은 설계단계 및 공용중 안정적인 응답을 확보하기 위해 필수적이다. 이 논문에서는 실축척 구조물을 대상으로 하는 강제 진동실험에서 일반적으로 사용할 수 있는 가진방법인 가진기 및 충격해머를 사용한 가진실험의 결과를 상호 비교하여 가진방법에 따른 고유진동수 추출결과의 차이 및 차이의 발생원인을 분석하고, 보다 정확한 구조물의 고유진동수를 산출하기 위한 보정방법을 제안하였다. 또한, 철도교량의 시공단계에 따라 궤도부설 전, 후에 진동실험을 수행하였으며, 궤도부설에 의한 고유진동수 변화를 관찰하였다. 가진방법에 따른 측정 고유진동수의 차이는 가진기의 경우 부가질량의 효과, 충격해머의 경우 동적응답 레벨에 따라 변화하는 지점강성효과에 기인하는 것으로 파악되었으며, 이에 대한 보정을 통해 보다 신뢰성 높은 고유진동수로 환산하는 것이 가능하였다. 궤도부설의 효과는 일반적으로 부가질량으로만 고려하여 궤도부설 전에 비해 고유진동수가 감소하는 것으로 예상되어 왔지만, 이 연구에서 수행한 실험의 결과로 질량추가의 효과에 상당하는 강성기여 효과도 있는 것으로 나타났으며, 강성증가율은 발생변위의 수준에 비례하여 증가하는 것으로 나타났다. 따라서, 설계단계에서 완성계의 고유진동수 예측시 궤도의 발생변위수준을 고려하여 적절히 강성기여 효과를 고려하는 것이 필요할 것으로 판단된다.

• Korean Journal of Radiology
• /
• 제20권8호
• /
• pp.1275-1284
• /
• 2019

Objective: To develop algorithms using convolutional neural networks (CNNs) for automatic segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) and compare them with conventional algorithms, including a thresholding-based segmentation. Materials and Methods: Between September 2005 and August 2015, 429 patients presenting with acute cerebral ischemia (training:validation:test set = 246:89:94) were retrospectively enrolled in this study, which was performed under Institutional Review Board approval. Ground truth segmentations for acute ischemic lesions on DWI were manually drawn under the consensus of two expert radiologists. CNN algorithms were developed using two-dimensional U-Net with squeeze-and-excitation blocks (U-Net) and a DenseNet with squeeze-and-excitation blocks (DenseNet) with squeeze-and-excitation operations for automatic segmentation of acute ischemic lesions on DWI. The CNN algorithms were compared with conventional algorithms based on DWI and the apparent diffusion coefficient (ADC) signal intensity. The performances of the algorithms were assessed using the Dice index with 5-fold cross-validation. The Dice indices were analyzed according to infarct volumes (< 10 mL, ≥ 10 mL), number of infarcts (≤ 5, 6-10, ≥ 11), and b-value of 1000 (b1000) signal intensities (< 50, 50-100, > 100), time intervals to DWI, and DWI protocols. Results: The CNN algorithms were significantly superior to conventional algorithms (p < 0.001). Dice indices for the CNN algorithms were 0.85 for U-Net and DenseNet and 0.86 for an ensemble of U-Net and DenseNet, while the indices were 0.58 for ADC-b1000 and b1000-ADC and 0.52 for the commercial ADC algorithm. The Dice indices for small and large lesions, respectively, were 0.81 and 0.88 with U-Net, 0.80 and 0.88 with DenseNet, and 0.82 and 0.89 with the ensemble of U-Net and DenseNet. The CNN algorithms showed significant differences in Dice indices according to infarct volumes (p < 0.001). Conclusion: The CNN algorithm for automatic segmentation of acute ischemic lesions on DWI achieved Dice indices greater than or equal to 0.85 and showed superior performance to conventional algorithms.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2007년도 추계학술대회논문집
• /
• pp.655-660
• /
• 2007

This paper presents real-time hybrid test method of large-scale MR damper applied to a building structure under seismic excitation. The real-time hybrid test using an actuator for the control performance evaluation of a MR damper controlling the response of earthquake-excited building structure is experimentally implemented. In the test, the building structure is used as a numerical part, on which a large-scale MR damper adopted as an experimental part was installed to reduce its response. At first, the force that is acting between a MR damper and building structure is measured from the load cell attached on the actuator system and is fed-back to the computer to control the motion of the actuator. Then, the actuator is so driven that the error between the interface displacement computed from the numerical building structure with the excitations of earthquake and the fed-back interface force and that measured from the actuator. The control efficiency of the MR damper used in this paper is experimentally confirmed by implementing this process of experiment on real-time.

• 전산구조공학
• /
• 제5권3호
• /
• pp.105-112
• /
• 1992

원자로 내부구조물의 설계시 필요한 동적응답해석을 위하여 각 구조물의 정확한 진동특성을 파악할 필요가 있다. 한국 표준형 원자력발전소를 위하여 설계된 제어봉집합체 보호구조물은 기존의 설계로 부터 많은 설계변경이 있었고, 또 이 구조물은 튜우브와 얇은 판이 사각격자 형태로 이루어져 있고 연결봉에 의해 고정되는 등 매우 복잡한 형태로 구성되어 있어서 해석과 시험에 의한 진동측정 프로그램을 수행할 필요성이 대두되었다. 따라서 본 논문에서는 진동측정 프로그램의 첫 단계로서 범용구조해석코드인 ANSYS를 이용하여 시험전 해석을 수행하였다. 또 자유도의 수와 얇은 판에 있는 구멍 및 연결봉의 pre-load가 구조물의 자유진동수에 미치는 영향을 검토하였다. 이로부터 결정된 유한요소모델에 대하여 모우드해석을 수행하여 구조물의 고유진동수와 모우드형상을 구하였고, 조화운동해석(Harmonic Analysis)을 행하여 주요모우드에 대한 응답을 측정함으로써 추후에 수행될 진동측정 시험조건 즉 응답측정부위, 측정위치의 수, 측정진동수의 범위 및 가진력의 크기 등을 결정하였다.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2003년도 춘계학술대회논문집
• /
• pp.905-910
• /
• 2003

One of the methods that are used to compare and verify the supporting performance of the spacer grids developed is the vibration characteristic test. A modal test in this paper is performed for a dummy rod 3,847mm tall supported by eight New Doublet (ND) spacer grids. For the vibration test in air, nine accelerometers, one displacement sensor and one shaker are used for acquiring signals, and an I-DEAS TDAS software is employed for analyzing the signals. Also, a finite element (FE) analysis is performed by a beam-spring simple model and a contact model simulating the contact phenomenon between the rod and the fm spring. And then, the result of the FE analysis is compared with that of the modal test. The natural frequencies as well as the mode shapes calculated by the proposed contact models have a greater similarity to the test results than those by the previous beam-spring model. In addition, for grasping whether or not the modal parameters are influenced by where shaking spot is, two kinds of tests are performed; one is for the shaker attached at the fourth span (center), the other is for the shaker at the fifth span that is one span nearer to the bottom of the rod. The latter shows higher MAC than the former. Finally, the vibration displacements are measured in the range of 0.112-0.214mm for the excitation force of 0.25-0.75 N.