• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.1-16
• /
• 2022

Despite recent breakthroughs in deep learning and computer vision fields, the pixel-wise identification of tiny objects in high-resolution images with complex disturbances remains challenging. This study proposes a modified U-net for tiny crack segmentation in real-world steel-box-girder bridges. The modified U-net adopts the common U-net framework and a novel Self-Attention-Self-Adaption (SASA) neuron as the fundamental computing element. The Self-Attention module applies softmax and gate operations to obtain the attention vector. It enables the neuron to focus on the most significant receptive fields when processing large-scale feature maps. The Self-Adaption module consists of a multiplayer perceptron subnet and achieves deeper feature extraction inside a single neuron. For data augmentation, a grid-based crack random elastic deformation (CRED) algorithm is designed to enrich the diversities and irregular shapes of distributed cracks. Grid-based uniform control nodes are first set on both input images and binary labels, random offsets are then employed on these control nodes, and bilinear interpolation is performed for the rest pixels. The proposed SASA neuron and CRED algorithm are simultaneously deployed to train the modified U-net. 200 raw images with a high resolution of 4928 × 3264 are collected, 160 for training and the rest 40 for the test. 512 × 512 patches are generated from the original images by a sliding window with an overlap of 256 as inputs. Results show that the average IoU between the recognized and ground-truth cracks reaches 0.409, which is 29.8% higher than the regular U-net. A five-fold cross-validation study is performed to verify that the proposed method is robust to different training and test images. Ablation experiments further demonstrate the effectiveness of the proposed SASA neuron and CRED algorithm. Promotions of the average IoU individually utilizing the SASA and CRED module add up to the final promotion of the full model, indicating that the SASA and CRED modules contribute to the different stages of model and data in the training process.

• Journal of the Korea Concrete Institute
• /
• v.24 no.6
• /
• pp.737-744
• /
• 2012

Hysteretic steel damper has been applied mainly to steel buildings. However, the usage in RC buildings is rapidly increasing recently. In order to apply the steel hysteretic damper in RC buildings, supporting elements of the damper should have sufficient strength and stiffness suitable for transferring damper forces to beams and walls. But due to the inevitable damage in reinforced concrete elements due to cracking, identification of the load transfer mechanism from damper to supporting element and hysteretic characteristics of the supporting element are extremely important in evaluating the damper behavior. Experiments were carried out on connection details between RC walls and supporting elements of the steel plate typed damper such as EaSy damper. The test results showed that fracture patterns of all specimens were almost identical except in the crack number and pattern associated with shear loading condition. Among the specimens, HD-3 shoed a well distributed cracks patterns along with good performance with respect to energy dissipation capacity, stiffness deterioration, and strength degradation.

• Smart Structures and Systems
• /
• v.22 no.3
• /
• pp.291-302
• /
• 2018

This paper presents a flexibility based method for damage identification from static measurements in beam-type structures. The response of the beam at the Damaged State is decomposed into the response at the Reference State plus the response at an Incremental State, which represents the effect of damage. The damage is localized by detecting slope discontinuities in the deflection of the structure at the Incremental State. A denoising filtering technique is applied to reduce the effect of experimental noise. The extent of the damage is estimated through comparing the experimental flexural stiffness of the damaged cross-sections with the corresponding values provided by analytical models of cracked beams. The paper illustrates the method by showing a numerical example with two cracks and an experimental case study of a simply supported steel beam with one artificially introduced notch type crack at three damage levels. A Digital Image Correlation system was used to accurately measure the deflections of the beam at a dense measurement grid under a set of point loads. The results indicate that the method can successfully detect and quantify a small damage from the experimental data.

• Journal of the Korea Concrete Institute
• /
• v.11 no.4
• /
• pp.107-116
• /
• 1999

The damage in concrete structures generally starts with microcracking and thus it is important to find and explore these microcracks in concrete in order to ensure appropriate safety and serviceability. The purpose of the present study is to identify the damage characteristics of concrete structures due to cracking by employing the acoustic emission techniques. A comprehensive experimental study has been done. The cracking damage under tensile and flexural loadings have been identified and the bond damage between steel and concrete have been also characterized. It is seen that the amplitudes and energy level of Acoustic Emission(AE) events are found to be relatively small for bond cracking damages and large for tensile cracking damages. The characteristic equations of the AE events for various cracking damages have been proposed based on the present test data. The internal microcracks are progressively developed ahead of a visible actual crack and the present study clearly exhibits these damage mechanism for various types of cracking in concrete. The present study provides useful data which can be used to identify the various types of cracking damages in concrete structures. This will allow efficient maintenance of concrete structures through monitoring of internal cracking based on acoustic emission.

• Proceedings of the Korea Concrete Institute Conference
• /
• 1999.04a
• /
• pp.543-546
• /
• 1999

The purpose of the present study is to identify the damage characteristics of concrete structures due to cracking by employing the acoustic emission techniques. A comprehensive experimental study has been done. The cracking damages under tensile and flexural loadings have been identified and the bond damage between steel and concrete have been also characterized. It is seen that the amplitudes and energy level of AE events is found to be smaller for bond cracking damages and larger for tensile cracking damages. The characteristic equations of the AE events for various cracking damages have been proposed based on the present test data. The internal microcracks are progressively developed ahead of a visible actual crack and the present study clearly exhibits thses damage mechanism for various types of cracking in concrete. The present study provides very useful data which can be used to identify the various types of cracking damages in concrete structures. This will allow very efficient maintenance of concrete structures through monitoring of internal cracking based on acoustic emission.

• Journal of the Korean Society for Precision Engineering
• /
• v.6 no.3
• /
• pp.90-99
• /
• 1989

In this study, J-integral values obtained by various methods, i.e, R-Curve, Unloading Compliance, Strectched Zone Width [SZW], and Acoustic Emission [AE] methods are investigated. Elasto-plastic fracture toughness [$J_IC$] estimations by R-curve method are overestimated than those by SZW method, and those by unloading compliance method is around middle value of them. And the difference between them is little. The $J_IC$ value by AE method was almost agreed with that by SZW, and then proved to be useful. Crack propagation mechanism on fractography is a stable ductile fracture. For the identification of ductile fracture, both fractography analysis and AE method were applied to estimate the characteristics more precisely.