• Korean Journal of Materials Research
• /
• v.3 no.2
• /
• pp.197-204
• /
• 1993

Abstract The effects of the hole size and the specimen width on the fracture behavior of several fabric composite plates are experimentally investigated in tension. Tests are performed on plain woven glass/ epoxy, plain woven carbon/epoxy and satin woven glass/polyester specimens with a circular hole. It is shown in this paper that the characteristic length according to the point stress criterion depends on the hole size and the specimen width. An excellent agreement is found between the experimental results and the analytical predictions of the modified failure criterion. The notched strength increase with an increase in the damage ratio, which is explained by a stress relaxation due to the formation of damage zone. When the unstable fracture occurred, the critical crack length equivalent for the damage zone is about twice the characteristic length. The critical energy release rate $G_c$ is independent of hole size for the same specimen width. The variation of $G_c$ according to the material system, fiber volume fraction and specimen width relates to the notch sensitivity factor. $G_c$ increases with a decrease in the notch sensitivity factor, which can be explained by a stress relaxation due to the increase of damage zone.

• Transactions of Materials Processing
• /
• v.25 no.6
• /
• pp.379-389
• /
• 2016

As a stator for steam turbine diaphragm, hollow-type nozzle stator to substitute for conventional solid one is introduced in this study. This hollowed stator can be separated into two parts such as upper and lower plates with large and curved surface area. This study focuses on thick plate forming process for the upper plate of the hollow-partitioned nozzle stator. First, to reduce forming defects such as under-cut and localized thinning of the deformed plate, and to avoid tool interruption between forming punch and lower die, tool design including the position determination of forming surfaces is performed. Uni-axial tensile tests are carried out using SUS409L steel plate with initial thickness of 5.00mm, and plastic strain ratio (r-value) is also obtained. Due to the asymmetric curved configuration of the upper plate, it is hard to adopt a series of blank holder or draw-bead, so the initial plate during this thick plate forming experiences unstable and non-uniform contact. To easy this forming difficulty and find suitable tool geometry without sliding behavior of the workpiece in the die cavity, two geometric parameters with respect to each shoulder angle of the lower die and the upper punch are adopted. FE models with consideration of 21 combinations for the geometric parameters are built-up, and numerical simulations are performed. From the simulated and predicted results, it is shown that the geometric parameter combinations with ($30^{\circ}$, $90^{\circ}$) and ($45^{\circ}$, $90^{\circ}$) for the shoulder angle of the lower die and the upper punch are suitably applied to this upper plate forming of the hollow-partitioned nozzle stator used for the turbine diaphragm.

• Geomechanics and Engineering
• /
• v.14 no.2
• /
• pp.195-202
• /
• 2018

Many laboratory experiments on crack propagation under uniaxial loading and biaxial loading have been conducted in the past using transparent materials such as resin, polymethyl methacrylate (PMMA), etc. However, propagation behaviors of three-dimensional (3D) cracks in rock or rock-like materials under tri-axial loading are often considerably different. In this study, a series of true tri-axial loading tests on the rock-like material with two semi-ellipse pre-existing cracks were performed in laboratory to investigate the acoustic emission (AE) characteristics and propagation characteristics of 3D crack groups influenced by intermediate principal stress. Compared with previous experiments under uniaxial loading and biaxial loading, the tests under true tri-axial loading showed that shear cracks, anti-wing cracks and secondary cracks were the main failure mechanisms, and the initiation and propagation of tensile cracks were limited. Shear cracks propagated in the direction parallel to pre-existing crack plane. With the increase of intermediate principal stress, the critical stress of crack initiation increased gradually, and secondary shear cracks may no longer coalesce in the rock bridge. Crack aperture decreased with the increase of intermediate principal stress, and the failure is dominated by shear fracturing. There are two stages of fracture development: stable propagation stage and unstable failure stage. The AE events occurred in a zone parallel to pre-existing crack plane, and the AE zone increased gradually with the increase of intermediate principal stress, eventually forming obvious shear rupture planes. This shows that shear cracks initiated and propagated in the pre-existing crack direction, forming a shear rupture plane inside the specimens. The paths of fracturing inside the specimens were observed using the Computerized Tomography (CT) scanning and reconstruction.

• Journal of Aerospace System Engineering
• /
• v.13 no.5
• /
• pp.32-38
• /
• 2019

The flying wing configuration with high sweep angles and rounded leading edge represent a complex flow of structures by the leading edge vortex. For control of the tailless flying wing configuration with unstable directional stability, flaperon is used. In this study, we conducted numerical simulations for a non-slender flying wing configuration with a rounded leading edge and analyzed the effect of the sideslip angle and flaperon. Through aerodynamic coefficient analysis, it was found that the effect of AoS on lift and drag coefficient was minimal and the side force and moment coefficient were markedly influenced by AoS. As the sideslip angle increased, the pitch break, which is related to the pitching moment coefficient, was delayed. Through stability analysis, the directional and lateral static stability of the flying wing configuration were increased by flaperon. Also, the structure and behavior of the leading edge vortex were analyzed by observing the contour of the pressure coefficient and the skin friction line.

• Composites Research
• /
• v.34 no.5
• /
• pp.323-329
• /
• 2021

In this paper, the characteristics of fracture in mode I loading were analyzed for adhesively bonded joints with non-uniform adhesion. The Double Cantilever Beam test was performed and mode I fracture toughness was obtained. In the case of non-uniform adhesively bonded joints, the stable crack growth sections and unstable crack growth section were shown. The fracture characteristics of each section were observed through the load-displacement curve of the DCB test and the fracture surface of the specimen. Finite Element Analysis was performed at the section based on segmented section by crack length measured through the test and using the mode I fracture toughness of each section. Through DCB test results and finite element analysis results, it was confirmed that the fracture behavior of specimens with non-uniform adhesion can be simulated.

• Herbal Formula Science
• /
• v.27 no.2
• /
• pp.121-136
• /
• 2019

Objective : To investigate the effect of sihogayonggolmoryeotang (SY) on Single Prolonged Stress(SPS)-induced Post Traumatic Stress Disorder(PTSD). Method : To confirm the effects of SY on SPS-induced PTSD, Changes in body weight, sucrose intake open field test(OFT) and forced swimming test(FST)were observed. After behavioral tests, the plasma corticosterone(CORT) from the abdominal aorta, serotonin(5-HT) from prefrontal cortex, hippocampus, amygdala and striatum, norepinephrine(NE) and dopamine(DA) from hippocampus was measured by ELISA. mRNA expression of brain-derived neurotrophic factor(BDNF) and cAMP response element-binding protein(CREB) in hippocampus was measured by RT-PCR. Result : Weight change and sucrose intakes of rats in 14th day after the administration of SY were significantly increased in the SPS + SY450 group compared to the SPS group (p<0.05). Numbers of crossing in the central zone in the OFT were significantly increased in the SPS + SY450 group (p<0.05) compared with the SPS group. The immobility time of FST was significantly decreased in SPS + SY450 group compared with SPS group (p<0.05). The change of plasma CORT concentration was significantly decreased in SPS + SY450 group compared with that in SPS group (p<0.05). The change of 5-HT concentration was significantly increased in the SPS + SY450 group at hippocampus and amygdala compared with the SPS group (p<0.05). The concentration of DA was significantly increased in the SPS + SY450 group compared with the SPS group (p<0.05). The expression of BDNF and CREB were significantly increased in SPS + SY450 group compared with the SPS group (p<0.05). Conclusion : SY administration lowered the increase of CORT caused by PTSD and increases the 5-HT concentration and reversed the decreased expression of NE and DA and BDNF and CREB by PTSD. It is postulated that SY is effective in treating PTSD by restoring cognitive function, memory impairment, unstable emotional disturbances.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.4
• /
• pp.13-27
• /
• 2021

In this study, the impact of traffic flow on the market penetration rate of autonomous vehicles(AV) was analyzed using the data for the year 2020 of the Yongin IC~Yangji IC section of Yeongdong Expressway. For this analysis, a microscopic traffic simulation model VISSIM was utilized. To construct the longitudinal control of the AV, the Intelligent Driver Model(IDM) was built and applied, and the driving behavior was verified by comparison with a normal vehicle. An examination of the study results of mobility and safety according to the market penetration rate of the AV, showed that the network's mobility improves as the market penetration rate increases. However, from the point of view of safety, the network becomes unstable when normal vehicles and AVs are mixed, so there should be a focus on traffic management for ensuring safety in mixed traffic situations.

• Journal of the Society of Disaster Information
• /
• v.18 no.1
• /
• pp.154-162
• /
• 2022

Purpose: The government intends to come up with measures to revitalize safety management at construction sites to shift safety management at construction sites from human capabilities to system-oriented management systems using advanced technologies AIoT and Mobile technologies. Method: The construction site safety management monitoring system using AIoT and Mobile technology conducted an experiment on the effectiveness of the construction site by applying three algorithms: virtual fence, fire monitoring, and recognition of not wearing a safety helmet. Result: The number of workers in the experiment was 215 and 7.61 virtual fence intrusion was 3.5% compared to the number of subjects and 0.16 fire detection were 0.07% compared to the subjects, and the average monthly rate of not wearing a safety helmet was 8.79, 4.05% compared to the subjects. Conclusion: It was found that the construction site safety management monitoring system using AIoT and Mobile technology has a valid effect on the construction site.

• Geomechanics and Engineering
• /
• v.37 no.1
• /
• pp.49-64
• /
• 2024

Rocks undergoing repeated loading and unloading over an extended period, such as due to earthquakes, human excavation, and blasting, may result in the gradual accumulation of stress and deformation within the rock mass, eventually reaching an unstable state. In this study, a CNN-CCM is proposed to address the mechanical behavior. The structure and hyperparameters of CNN-CCM include Conv2D layers × 5; Max pooling2D layers × 4; Dense layers × 4; learning rate=0.001; Epoch=50; Batch size=64; Dropout=0.5. Training and validation data for deep learning include 71 rock samples and 122,152 data points. The AI Rock Constitutive Model learned by CNN-CCM can predict strain values(ε₁) using Mass (M), Axial stress (σ₁), Density (ρ), Cyclic number (N), Confining pressure (σ₃), and Young's modulus (E). Five evaluation indicators R², MAPE, RMSE, MSE, and MAE yield respective values of 0.929, 16.44%, 0.954, 0.913, and 0.542, illustrating good predictive performance and generalization ability of model. Finally, interpreting the AI Rock Constitutive Model using the SHAP explaining method reveals that feature importance follows the order N > M > σ₁ > E > ρ > σ₃.Positive SHAP values indicate positive effects on predicting strain ε₁ for N, M, σ₁, and σ₃, while negative SHAP values have negative effects. For E, a positive value has a negative effect on predicting strain ε₁, consistent with the influence patterns of conventional physical rock constitutive equations. The present study offers a novel approach to the investigation of the mechanical constitutive model of rocks under cyclic loading and unloading conditions.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.