• Journal of the Korean Geotechnical Society
• /
• v.37 no.5
• /
• pp.5-17
• /
• 2021

In this study, the shear behavior between pile-sandy soil interface was quantified based on series of rigorous ring shear test results. Ring shearing test was carried out to observe the shear behavior prior to failure and behavior at residual state between most commonly used pile materials - steel and concrete - and Jumunjin sand. The test was set to clarify the shear behavior under various confinement conditions and soil densities. The test results were converted in to representative friction angles for various test materials. Additional numerical analysis was executed to validate the accuracy of the test results. Based on the test results and the numerical validation, it was found that due to the dilative and contractive nature of sand, its interface behavior can be categorized in to two different types : soils with higher densities tend to show peak shear stress and moves on to residual state, while on the other hand, soils with lower densities tend to show bilinear load-transfer curves along the interface. However, the relative density and the confining stress was found to affect the friction angle only in the small train range, and converges as it progresses to large deformation. This study established a large deformation analysis method which can successfully simulate and predict the large deformation behavior such as ring shear tests. Moreover, the friction angle derived from the ring shear test result and verified by numerical analysis can be applied to numerical analysis and actual design of various pile foundations.

• 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 Computational Structural Engineering Institute of Korea
• /
• v.36 no.3
• /
• pp.185-192
• /
• 2023

In this study, fluctuating wind velocity for time history analysis is simulated by a single variate, single-dimensional random process using the KBC2022 spectrum about across-wind direction. This study analyzed and obtained the inelastic dynamic response for structures modeled as a single-degree-of-freedom system. It is assumed that the wind response is excellent in the primary mode, the change in vibration owing to plasticization is minor, along-wind vibration and across-wind vibration are independent, and the effect of torsional vibration is small. The numerical results, obtained by the Newmark-𝛽 method, shows the time-history responses and trends of maximum displacements. As a result of analyzing the inelastic dynamic response of the structure with the second stiffness ratio(𝛼) and yield displacement ratio (𝛽) as variables, it is identified that as the yield displacement ratio (𝛽) increases when the second stiffness ratio is constant, the maximum displacement ratio decreases, then reaches a minimum value, and then increases. When the stiffness ratio is greater than 0.5, there is a yield point ratio at which the maximum displacement ratio is less than 1, indicating that the maximum deformation is reduced compared to the elastically designed building even if the inelastic behavior is permitted in the inelastic wind design.