• 한국지진공학회논문집
• /
• 제8권6호통권40호
• /
• pp.73-80
• /
• 2004

구조물의 손상 진단을 위해 측정 가속도 데이터를 사용한 비선형 시간영역 SI 알고리듬을 개발하였다. 구조물의 비선형 거동을 고려하기 위하여 측정 가속도 증분과 해석에 의한 가속도 증분의 차이로 출력오차를 정의하고, 구속 비선형 최적화 문제를 풀어 최적 구조변수를 구하였다. 개발된 알고리듬은 시간에 따라 변하는 강성도와 감쇠 변수를 추정하도록 하였다. 구조물의 비선형 거동에 의한 복원력은 추정된 시간에 따라 변하는 구조변수와 Newmark-$\beta$법으로 계산한 변위를 사용하여 복원하였으며, 복원 과정에서 비탄성 거동에 대한 어떤 모델도 사전에 설정하지 않았다. 개발한 알고리듬에서는 측정오차와 공간 및 상태에 대한 불완전 측정의 경우를 고려하였다. 개발한 알고리듬을 검증하기 위하여 3층 전단건물에 대한 수치 모의시험과 실내 모형실험을 통한 연구를 수행하였다.

• 한국전산구조공학회논문집
• /
• 제32권4호
• /
• pp.205-213
• /
• 2019

2017년 발생한 포항 지진으로 인하여 천장재, 외장재, 커튼월 등 비구조재의 파괴에 의한 피해가 다수 보고되었으며 비구조재의 내진설계가 중요해지고 있다. 본 연구에서는 임팩트해머 테스트를 통해 행어볼트 길이에 따른 천장재의 고유진동수와 감쇠비를 식별하였다. 또한 천장재가 벽 또는 다른 구조체에 충돌하는 경우 발생하는 충격효과를 정확히 고려하기 위해 충돌실험을 수행하였다. 식별된 천장재의 동특성과 충격지속시간을 바탕으로 실제로 천장재가 지진하중으로 인하여 주변 구조물과 충돌이 발생하는 경우에 대한 천장재 응답특성을 수치해석을 통하여 분석하였다. 수치해석 시뮬레이션 결과, 충격하중은 이격거리에 따라 선형적으로 증가하는 경향을 보였으며, 달대길이와는 무관한 것으로 나타났다.

• 한국전산구조공학회논문집
• /
• 제33권3호
• /
• pp.193-199
• /
• 2020

본 논문은 유한요소법과 유전알고리즘을 연동하여 지진하중을 받는 구조물의 강성저하(손상) 및 보강 후 효과를 추정하는 방법을 다루었다. 본 연구의 독창성은 지진하중을 적용하였고, 그 응답으로부터 구조물의 미지 변수를 추정한다는 점이다. 본 연구에서 제안한 방법은 지진하중으로부터 손상된 부위를 추정할 뿐 아니라, 그 위치와 정도를 규명할 수 있다. 제안한 방법을 검증하기 위하여 El Centro 및 포항 지진하중을 적용하여 저층 뼈대구조물와 트러스 교량을 대상으로 알고리즘을 실행하였다. 수치해석 예제는 제안한 방법이 수치해석적인 효율성 뿐 아니라 지진으로부터의 심각한 피해를 예방하는 데 적용할 수 있음을 보여주었다.

• Computers and Concrete
• /
• 제33권4호
• /
• pp.341-348
• /
• 2024

Ultra-high-performance concrete (UHPC) has received remarkable attentions in civil infrastructure due to its unique mechanical characteristics and durability. UHPC gains increasingly dominant in essential structural elements, while its unique properties pose challenges for traditional inspection methods, as damage may not always manifest visibly on the surface. As such, the need for robust inspection techniques for detecting cracks in UHPC members has become imperative as traditional methods often fall short in providing comprehensive and timely evaluations. In the era of artificial intelligence, computer vision has gained considerable interest as a powerful tool to enhance infrastructure condition assessment with image and video data collected from sensors, cameras, and unmanned aerial vehicles. This paper presents a computer vision-based approach employing deep learning to detect cracks in UHPC beams, with the aim of addressing the inherent limitations of traditional inspection methods. This work leverages computer vision to discern intricate patterns and anomalies. Particularly, a convolutional neural network architecture employing transfer learning is adopted to identify the presence of cracks in the beams. The proposed approach is evaluated with image data collected from full-scale experiments conducted on UHPC beams subjected to flexural and shear loadings. The results of this study indicate the applicability of computer vision and deep learning as intelligent methods to detect major and minor cracks and recognize various damage mechanisms in UHPC members with better efficiency compared to conventional monitoring methods. Findings from this work pave the way for the development of autonomous infrastructure health monitoring and condition assessment, ensuring early detection in response to evolving structural challenges. By leveraging computer vision, this paper contributes to usher in a new era of effectiveness in autonomous crack detection, enhancing the resilience and sustainability of UHPC civil infrastructure.

• 한국전산구조공학회논문집
• /
• 제26권2호
• /
• pp.99-112
• /
• 2013

이 논문은 부분공간 시스템 확인기법을 이용하여 전단빌딩의 강성행렬과 부재의 강성을 추정하는 기법을 소개한다. 시스템 행렬은 입력-출력 데이터로 구성된 행켈행렬을 LQ 분해와 특이치 분해를 통해 추정한다. 추정된 시스템 행렬은 닮음 변환을 통해 실제 좌표축으로 변환하고, 변환된 시스템 행렬로부터 강성행렬을 계산한다. 추정된 강성행렬의 정확성과 안정성은 행켈행렬의 크기에 따라 변한다. 전단빌딩의 기저 유한요소 모델을 이용하여 행켈행렬의 크기에 따른 강성행렬의 추정 오차 곡선을 구한다. 오차 곡선을 이용하여 목표 정확도 수준에 부합하는 행켈행렬의 크기들을 결정한다. 이렇게 선택된 행렬의 크기들 중에서 부분공간 시스템 확인의 계산비용을 고려하여 보다 적절한 행렬의 크기를 결정할 수 있다. 결정된 크기의 행켈행렬을 이용하여 강성행렬을 추정하고 추정된 강성행렬로부터 부재의 강성을 추정한다. 제안된 방법을 손상 전후의 5층 전단빌딩 수치 예제에 적용하여 타당성을 검증한다.

• Earthquakes and Structures
• /
• 제11권4호
• /
• pp.629-648
• /
• 2016

This paper deals with 111 buildings built between 1962 and 1987, from various parts of the city of Osijek, for which, through the collection of documentation, a database is created. The aim of this paper is to provide the first steps in assessing seismic risk in Osijek applying method based on vulnerability index. This index uses collected information of parameters of the building: the structural system, the construction year, plan, the height, i.e., the number of stories, the type of foundation, the structural and non-structural elements, the type and the quality of main construction material, the position in the block and built-up area. According to this method defining five damage states, the action is expressed in terms of the macroseismic intensity and the seismic quality of the buildings by means of a vulnerability index. The value of the vulnerability index can be changed depending on the structural systems, quality of construction, etc., by introducing behavior and regional modifiers based on expert judgments. Since there is no available data of damaged buildings under earthquake loading in our country, we will propose behavior modifiers based on values suggested by earlier works and on judgment based on available project documentation of the considered buildings. Depending on the proposed modifiers, the seismic vulnerability of existing buildings in the city of Osijek will be assessed. The resulting vulnerability of the considered residential buildings provides necessary insight for emergency planning and for identification of critical objects vulnerable to seismic loading.

• Smart Structures and Systems
• /
• 제29권1호
• /
• pp.129-140
• /
• 2022

The effectiveness of system identification, damage detection, condition assessment and other structural analyses relies heavily on the accuracy and reliability of the measured data in structural health monitoring (SHM) systems. However, data anomalies often occur in SHM systems, leading to inaccurate and untrustworthy analysis results. Therefore, anomalies in the raw data should be detected and cleansed before further analysis. Previous studies on data anomaly detection mainly focused on just single type of data anomaly for denoising or removing outliers, meanwhile, the existing methods of detecting multiple data anomalies are usually time consuming. For these reasons, recognising multiple anomaly patterns for real-time alarm and analysis in field monitoring remains a challenge. Aiming to achieve an efficient and accurate detection for multi-type data anomalies for field SHM, this study proposes a pattern-recognition-based data anomaly detection method that mainly consists of three steps: the feature extraction from the long time-series data samples, the training of a pattern recognition neural network (PRNN) using the features and finally the detection of data anomalies. The feature extraction step remarkably reduces the time cost of the network training, making the detection process very fast. The performance of the proposed method is verified on the basis of the SHM data of two practical long-span bridges. Results indicate that the proposed method recognises multiple data anomalies with very high accuracy and low calculation cost, demonstrating its applicability in field monitoring.

• Computers and Concrete
• /
• 제22권2호
• /
• pp.227-237
• /
• 2018

Today, the modeling of concrete as a material within finite element simulations is predominantly done through nonlinear material models of concrete. In current sophisticated computational systems, there are a number of complex concrete material models which are based on theory of plasticity, damage mechanics, linear or nonlinear fracture mechanics or combinations of those theories. These models often include very complex constitutive relations which are suitable for the modeling of practically any continuum mechanics tasks. However, the usability of these models is very often limited by their parameters, whose values must be defined for the proper realization of appropriate constitutive relations. Determination of the material parameter values is very complicated in most material models. This is mainly due to the non-physical nature of most parameters, and also the large number of them that are frequently involved. In such cases, the designer cannot make practical use of the models without having to employ the complex inverse parameter identification process. In continuum mechanics, however, there are also constitutive relations that require the definition of a relatively small number of parameters which are predominantly of a physical nature and which describe the behavior of concrete very well within a particular task. This paper presents an example of such constitutive relations which have the potential for implementation and application in finite element systems. Specifically, constitutive relations for modeling the plane stress state of concrete are presented and subsequently tested and evaluated in this paper. The relations are based on the incremental theory of elastic strain-hardening plasticity in which a non-associated flow rule is used. The calculation result for the case of concrete under uniaxial compression is compared with the experimental data for the purpose of the validation of the constitutive relations used.

• Smart Structures and Systems
• /
• 제29권1호
• /
• pp.181-193
• /
• 2022

Data-driven structural health monitoring (SHM) of civil infrastructure can be used to continuously assess the state of a structure, allowing preemptive safety measures to be carried out. Long-term monitoring of large-scale civil infrastructure often involves data-collection using a network of numerous sensors of various types. Malfunctioning sensors in the network are common, which can disrupt the condition assessment and even lead to false-negative indications of damage. The overwhelming size of the data collected renders manual approaches to ensure data quality intractable. The task of detecting and classifying an anomaly in the raw data is non-trivial. We propose an approach to automate this task, improving upon the previously developed technique of image-based pre-processing on one-dimensional (1D) data by enriching the features of the neural network input data with multiple channels. In particular, feature engineering is employed to convert the measured time histories into a 3-channel image comprised of (i) the time history, (ii) the spectrogram, and (iii) the probability density function representation of the signal. To demonstrate this approach, a CNN model is designed and trained on a dataset consisting of acceleration records of sensors installed on a long-span bridge, with the goal of fault detection and classification. The effect of imbalance in anomaly patterns observed is studied to better account for unseen test cases. The proposed framework achieves high overall accuracy and recall even when tested on an unseen dataset that is much larger than the samples used for training, offering a viable solution for implementation on full-scale structures where limited labeled-training data is available.

• 한국강구조학회 논문집
• /
• 제17권1호통권74호
• /
• pp.1-11
• /
• 2005

이 논문에서는 구조물의 동적응답을 입력으로 하고, 패턴인식을 위해 신경망기법(Neural Network, NN)을 사용하는 손상감지기법을 제시하였다. 입력된 동적응답, 즉 주파수응답함수(FRF) 또는 변형률 주파수응답함수(SFRF)의 변화를 정량적으로 표현하기 위해 신호변형지수(Signal Anomaly Index, SAI)를 고안하여 사용하였으며, 이 신호변형지수는 손상 전 및 손상 후의 구조물로부터 측정된 가속도 또는 동적 변형률 신호를 사용하여 계산된다. 제안된 알고리즘은 2단계로 구성되며, 1단계에서는 신호변형지수 값의 크기 변화를 사용하여 구조물의 손상발생 유무를 판별하고, 여기서 구조물에 손상이 발생한 것으로 분석되면 2단계에서 신경망기법을 사용한 패턴인식을 통해 손상의 위치를 찾아낸다. 이 방법의 타당성 및 적용성을 확인하기 위해 강교량 축소모형에 대한 실험을 수행하였다. 신경망의 학습에는 수치해석을 통해 생성한 가상 신호를 사용하였으며, 학습이 완료된 신경망과 실험을 통해 측정한 실제 신호를 사용하여 손상발견을 수행하였다. 모형 교량에 대한 적용 결과로부터 이 알고리즘의 타당성이 검증되었으며, 향후 실 교량에 대한 적용도 가능할 것으로 판단된다.