• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.31 no.2
• /
• pp.79-85
• /
• 2018

A two-way pump can reduce costs by draining and circulating water out and into the drum of the washing machine using a single motor whereas a conventional one-way pump uses two motors for doing the same function. However, when the water is drained through the drainage outlet in the two-way pump casing, a backward or inhalation flow occurs and the water flows to the circulation outlet. Likewise, when the water is circulated, the backward flow or inhalation makes the water flow to the drainage outlet. In this study, design optimization of the two-way pump casing is performed to maximize its performance while improving backward flow and inhalation occurring inside of the pump casing. For this, design variables of the pump casing that mainly affect the performance of the pump such as flow rate and torque of the motor were selected through the analysis of mean. Using response surface models for the performances, the ratio of the flow rate to the torque was maximized with satisfying the constraints for the back flow and inhalation through design optimization.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.17 no.4
• /
• pp.375-387
• /
• 2004

A new finite element model will be presented to analyze the nonlinear behavior of RC beams and slabs strengthened by a patch repair. The numerical approach is based on the p-version degenerate shell element including theory of anisotropic laminated composites, theory of materially and geometrically nonlinear plates. In the nonlinear formulation of this model, the total Lagrangian formulation is adopted with large deflections and moderate rotations being accounted for in the sense of von Karman hypothesis. The material model is based on hardening rule, crushing condition, plate-end debonding strength model and so on. The Gauss-Lobatto numerical quadrature is applied to calculate the stresses at the nodal points instead of Gauss points. The validity of the proposed p-version nonlinear finite element model is demonstrated through the load-deflection curves, the ultimate loads, and the failure modes of RC beams or slabs bonded with steel plates or FRP plates compared with available result of experiment and other numerical methods.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.30 no.2
• /
• pp.119-125
• /
• 2017

Ablative material in a rocket nozzle is exposed to high temperature combustion gas, thus undergoes complicated thermal/chemical change in terms of chemical destruction of surface and thermal decomposition of inner material. Therefore, method for conjugate analysis of thermal response inside carbon/phenolic material including rocket nozzle flow, surface chemical reaction and thermal decomposition is developed in this research. CFD is used to simulate flow field inside nozzle and conduction in the ablative material. A change in material density and a heat absorption caused by the thermal decomposition is considered in solid energy equation. And algebraic equation under boundary layer assumption is used to deduce reaction rate on the surface and resulting destruction of the surface. In order to test the developed method, small rocket nozzle is solved numerically. Although the ablation of nozzle throat is deduced to be higher than the experiment, shape change and temperature distribution inside material is well predicted. Error in temperature with experimental results in rapid heating region is found to be within 100 K.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.33 no.5
• /
• pp.311-318
• /
• 2020

This study introduces a technique for improving the elastomeric-isolator performance using modular devices. The modular devices are shear resistance block, polymer spring, displacement acceptance guide, and anti-falling block. They are installed on the elastomeric isolator as a supplementary device. Each modularized device improves the isolator performance by performing step-by-step actions according to the seismic intensity and displacement. The PRB isolation device works in four stages, depending on the seismic magnitude, to satisfy the target performance. It is designed to accommodate design displacement in the first stage and large magnitude of earthquakes in the second and third stages. This design prevents superstructures from falling in the fourth stage due to large-magnitude earthquakes by increasing the capacity limit of the elastomeric isolator. In this study, the PRB isolation device is analyzed using finite element analysis to verify that the PRB isolation device works as intended and it can withstand loads corresponding to large-magnitude earthquakes. The performance of the PRB isolation device is validated by the analysis, which is further corroborated by actual experiments.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.5
• /
• pp.169-176
• /
• 2020

Recently, deep neural networks (DNNs) have brought revolutions to many mobile and embedded devices by providing human-level machine intelligence for various applications. However, high inference accuracy of such DNNs comes at high computational costs, and, hence, there have been significant efforts to reduce computational overheads of DNNs either by compressing off-the-shelf models or by designing a new small footprint DNN architecture tailored to resource constrained devices. One notable recent paradigm in designing small footprint DNN models is sharing parameters in several layers. However, in previous approaches, the parameter-sharing techniques have been applied to large deep networks, such as ResNet, that are known to have high redundancy. In this paper, we propose a parameter-sharing method for already parameter-efficient small networks such as ShuffleNetV2. In our approach, small templates are combined with small layer-specific parameters to generate weights. Our experiment results on ImageNet and CIFAR100 datasets show that our approach can reduce the size of parameters by 15%-35% of ShuffleNetV2 while achieving smaller drops in accuracies compared to previous parameter-sharing and pruning approaches. We further show that the proposed approach is efficient in terms of latency and energy consumption on modern embedded devices.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.29 no.6
• /
• pp.521-527
• /
• 2016

This paper is concerned with the numerical analysis of dynamic response of floating offshore wind turbine subject to underwater explosion using an effective non-reflecting technique. An infinite sea water domain was truncated into a finite domain, and the non-reflecting technique called the perfectly matched layer(PML) was applied to the boundary of truncated finite domain to absorb the inherent reflection of out-going impact wave at the boundary. The generalized transport equations that govern the inviscid compressible water flow was split into three PML equations by introducing the direction-wise absorption coefficients and state variables. The fluid-structure interaction problem that is composed of the wind turbine and the sea water flow was solved by the iterative coupled Eulerian FVM and Largangian FEM. And, the explosion-induced hydrodynamic pressure was calculated by JWL(Jones-Wilkins-Lee) equation of state. Through the numerical experiment, the hydrodynamic pressure and the structural dynamic response were investigated. It has been confirmed that the case using PML technique provides more reliable numerical results than the case without using PML technique.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.12 no.4
• /
• pp.525-535
• /
• 1999

In this study, a new method of fabricated concrete deck bridge construction is proposed. This paper details the method in which concrete multi-girders and fabricated concrete decks are rested on the upper flange of the girder and the female to female type sheat-key is formed to connect girder and deck. The finite element analysis is performed to verify the accuracy of the structural behaviors of the fabricated concrete deck bridge by comparing with experimental results. The first task performed is the analysis of the equilibrium of the member force occurring between the deck and the girder. After verifying equilibrium of the member force determined by the finite element analysis, this process is applied to the analysis of maximum member force as the position of design load. This task is utilized to determine the safety of each member according to the same scale finite element model. The final process in this study is to compare the deflection of girders used in experiment with that of the same scale finite element model to verify the strength of fabricated cincrete deck bridge. By this comparison, it is shown that the behavior of the fabricated concrete deck bridge is almost same as the finite element analysis. The second task is to analyze the load distribution effect according to the number of diaphragms and the composite effect due to the cinnection of the deck and girder by the finite element analysis. From the results of second task, it is found that the load distribution effect is not related to the number of diaphragms in case of the central loading, but is related to the number of diaphragms for eccentric loading. Analysis of the load distribution indicates that the effective number of diaphragm is three. It is also shown that the maximum deflection is decreased to almost one half due to the composite action of the deck and girder.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.21 no.1
• /
• pp.35-42
• /
• 2008

In the nuclear power plant, early detection of fatigue crack by non-destructive test (NDT) equipment due to the thermal cyclic load is very important in terms of strict safety regulation. To this end, many efforts are exerted to the fabrication of artificial cracked specimen for practicing engineers in the NDT company. The crack of this kind, however, cannot be made by conventional machining, but should be made under thermal cyclic load that is close to the in-situ condition, which takes tremendous time due to the repetition. In this study, thermal loading condition is investigated to minimize the time for fabricating the cracked specimen using simulation technique which predicts the crack initiation and propagation behavior. Simulation and experiment are conducted under an initial assumed condition for validation purpose. A number of simulations are conducted next under a variety of heating and cooling conditions, from which the best solution to achieve minimum time for crack with wanted size is found. In the simulation, general purpose software ANSYS is used for the stress analysis, MATLAB is used to compute crack initiation life, and ZENCRACK, which is special purpose software for crack growth prediction, is used to compute crack propagation life. As a result of the study, the time for the crack to reach the size of 1mm is predicted from the 418 hours at the initial condition to the 319 hours at the optimum condition, which is about 24% reduction.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.59 no.6
• /
• pp.61-71
• /
• 2017

The tracer gas method has an advantage that can estimate total and local ventilation rate by tracing air flow. However, the field measurement using tracer gas has disadvantages such as danger, inefficiency, and high cost. Therefore, the aim of this study was to evaluate ventilation rate in pig house by using the thermal distribution data rather than tracer gas. Especially, LMA (Local Mean Age), which is an index based on the age of air theory, was used to evaluate the ventilation rate in pig house. Firstly, the field experiment was conducted to measure micro-climate inside pig house, such as the air temperature, $CO_2$ concentration and wind velocity. And then, LMA was calculated based on the decay of $CO_2$ concentration and air temperature, respectively. This study compared between LMA determined by $CO_2$ concentration and air temperature; the average error and root mean square error were 3.76 s and 5.34 s. From these results, it was determined that thermal distribution data could be used for estimation of LMA. Finally, CFD (Computational fluid dynamic) model was validated using LMA and wind velocity. The mesh size was designed to be 0.1 m based on the grid independence test, and the Standard $k-{\omega}$ model was eventually chosen as the proper turbulence model. The developed CFD model was highly appropriate for evaluating the ventilation rate in pig house.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.12
• /
• pp.471-480
• /
• 2016

Recently, the damage of cyber attack toward infra-system, national defence and security system is gradually increasing. In this situation, military recognizes the importance of cyber warfare, and they establish a cyber system in preparation, regardless of the existence of threaten. Thus, the study of Intrusion Detection System(IDS) that plays an important role in network defence system is required. IDS is divided into misuse and anomaly detection methods. Recent studies attempt to combine those two methods to maximize advantagesand to minimize disadvantages both of misuse and anomaly. The combination is called Hybrid IDS. Previous studies would not be inappropriate for near real-time network environments because they have computational complexity problems. It leads to the need of the study considering the structure of IDS that have high detection rate and low computational cost. In this paper, we proposed a Hybrid IDS which combines C4.5 decision tree(misuse detection method) and Weighted K-means algorithm (anomaly detection method) hierarchically. It can detect malicious network packets effectively with low complexity by applying mutual information and genetic algorithm based efficient feature selection technique. Also we construct upgraded the the hierarchical structure of IDS reusing feature weights in anomaly detection section. It is validated that proposed Hybrid IDS ensures high detection accuracy (98.68%) and performance at experiment section.