• 기술발표회
• /
• s.2006
• /
• pp.320-328
• /
• 2006

In this test, there was two dimensional model loading test implemented for analysis with respect to the problem of evaluating bearing capacity and the application range on the dredged and reclaimed ground. It was got following conclusion through comparison of button's and Brown&Meyerhof"s equation with experimental result that was obtained by 2 dimensions model loading test. For the difference between average undrained shear strength by 2/3B of loading board width and under 2/3B is more than ${\pm}$ 50%, application of Nc(coefficient of bearing capacity was used in that case $\phi$=0 analysis is considered in the single layer) was declined. Brown&Meyerhof(1969)'s equation was underestimated comparing with loading test result, while Button(1953)'s equation was overestimated comparing with loading test result applied dividing as double layers of upper dessication layer and lower soft layer about dredged and reclaimed ground. Also, bearing capacity factors, Nc that was calculated by using button's equation was estimated greatly about 1.7 times more than bearing capacity factors, Nc that was calculated by using Brown&Meyerhof's equation. Bearing capacity factors, Nc that was calcuated by using Brown&Meyerhof's and Button's equation was evaluated each 2.3-3.6 times, 1.3-2.1 times smaller than bearing capacity factors, Nc5.14 that was calcuated by using Meyerhof's equation in case of unit layer.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.13 no.4
• /
• pp.780-788
• /
• 1989

An experimental study has been performed to investigate the process from nonequilibrium state to equilibrium state in redeveloping turbulent boundary layer beyond separation-reattachment using pitot tube and hot-wire anemometer. The model sued in the experiment has the form of a backward facing step which is assembled by a two-dimensional 4:1 half elipse and a plate. Measurements are carried out up to a distance of about 50 step height downstream of the step, where the reattachment observed at about x/h=6.5. The profiles of the shape factor H the Clauser parameter G and the coefficient of friction $C^{f}$ exhibited the characteristics similar to those of the equilibrium turbulent boundary layer from x/h=25, and the profiles of the trubulent quantities did from x/h=35. However, the wake region of the boundary layer does not seem to recover the equilibrium turbulent boundary layer even at x/h=50. By considering the distributions of the intermittency factor it has been noted that the turbulence structure changes gradually from a mixing layer to a turbulent boundary layer along downstream direction after reattachment. This becomes clearer as we analyse the one-dimensional energy spectra and the dissipation energy spectra which are measured and caculated at various downstream positions after the backward facing step.p.

• Journal of the Society of Naval Architects of Korea
• /
• v.59 no.1
• /
• pp.18-28
• /
• 2022

The turbulence stimulation effect of studs for boundary layer over a flat plate was investigated through the flow measurement in KRISO cavitation tunnel. For the test, Laser Doppler Velocimetry (LDV) and three flat plate models were used: (1) flat plate without studs; (2) flat plate with one stud row; (3) flat plate with two stud rows. The dimension and location of stud rows and the inflow speed were selected considering test conditions for standard-sized model ships in KRISO towing tank. The boundary layer characteristics of test models were analyzed and compared in terms of mean velocity profiles, turbulence intensity profiles, boundary layer thickness, and shape factor. In the case of the flat plate without studs, transition from laminar to turbulent flow occurred around Re_x=3.83 ~ 5.19 × 10⁵. In the case of flat plates with stud rows, the flow rapidly changed into turbulent flow right after passing the first stud row. In the state where turbulence was already developed, the second stud row slightly increased the turbulence intensity near the top of the stud, but did not significantly affect the boundary layer characteristics such as mean velocity distribution, boundary layer thickness, and shape factor.

• Asian Journal of Atmospheric Environment
• /
• v.10 no.2
• /
• pp.67-79
• /
• 2016

The forecasting of air pollution is an important and popular topic in environmental engineering. Due to health impacts caused by unacceptable particulate matter (PM) levels, it has become one of the greatest concerns in metropolitan cities like Karaj City in Iran. In this study, the concentration of $PM_{2.5}$ was predicted by applying a multilayer percepteron (MLP) neural network, a radial basis function (RBF) neural network and a Markov chain model. Two months of hourly data including temperature, NO, $NO_2$, $NO_x$, CO, $SO_2$ and $PM_{10}$ were used as inputs to the artificial neural networks. From 1,488 data, 1,300 of data was used to train the models and the rest of the data were applied to test the models. The results of using artificial neural networks indicated that the models performed well in predicting $PM_{2.5}$ concentrations. The application of a Markov chain described the probable occurrences of unhealthy hours. The MLP neural network with two hidden layers including 19 neurons in the first layer and 16 neurons in the second layer provided the best results. The coefficient of determination ($R^2$), Index of Agreement (IA) and Efficiency (E) between the observed and the predicted data using an MLP neural network were 0.92, 0.93 and 0.981, respectively. In the MLP neural network, the MBE was 0.0546 which indicates the adequacy of the model. In the RBF neural network, increasing the number of neurons to 1,488 caused the RMSE to decline from 7.88 to 0.00 and caused $R^2$ to reach 0.93. In the Markov chain model the absolute error was 0.014 which indicated an acceptable accuracy and precision. We concluded the probability of occurrence state duration and transition of $PM_{2.5}$ pollution is predictable using a Markov chain method.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.35 no.2
• /
• pp.169-178
• /
• 2011

The dispersion of a pollutant in a turbulent boundary layer has been described in this study by using a two-dimensional lattice Boltzmann method (LBM) and the Smagorinsky sub-grid-scale (SGS) model. The scalar transport equation corresponding to the pollutant concentration is adopted; the pollutant is considered to be in a continuous phase. The pollutant source is classified as ground-level source (GLS) and elevated-point source (ES). Air velocity and particle concentration profile for the pollutant are compared with the respective results and profiles obtained in the experiments of Fackrell and Robins (1982) and Raupach and Legg (1983). The numerical results obtained in this study, i.e., the simulation and the experimental data for the mean flow velocity profiles and the pollutant concentration profiles, are in good agreement with each other.

• The KIPS Transactions:PartD
• /
• v.9D no.6
• /
• pp.1137-1144
• /
• 2002

Possibility of federation improvement requires flexibility and adaptability of member federates. Moreover, to develop and convert a non-HLA(High Level Architecture) compliant model as a HLA federate and allow this federate to be integrated with a changeable federation, more time and efforts will be necessary. In this research, I proposed a method to design and implement a ROM (RTI Object Model) Framework as a solution to this problem. ROM completely separates RTI (Run-Time Infrastructure) programming and simulation programming therefore providing epochal efficiencies in cost and productivity to the development a HLA federate that supports a changeable FOM. That is, ROM contains two layers : 1) a management layer that manages RTI services between the RTI and the federate and 2) a Foundation Class layer that actually updates/reflects objects and interactions. These two layers allow federate developers to use more generalized HLA services and automates the iterative, low-level RTI programming process.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.2
• /
• pp.398-406
• /
• 2011

In this paper, we introduce a new architecture of PSO-based Polynomial Neural Networks (PNN) and discuss its comprehensive design methodology. The conventional PNN is based on a extended Group Method of Data Handling (GMDH) method, and utilized the polynomial order (viz. linear, quadratic, and modified quadratic) as well as the number of node inputs fixed (selected in advance by designer) at Polynomial Neurons located in each layer through a growth process of the network. Moreover it does not guarantee that the conventional PNN generated through learning results in the optimal network architecture. The PSO-based PNN results in a structurally optimized structure and comes with a higher level of flexibility that the one encountered in the conventional PNN. The PSO-based design procedure being applied at each layer of PNN leads to the selection of preferred PNs with specific local characteristics (such as the number of input variables, input variables, and the order of the polynomial) available within the PNN. In the sequel, two general optimization mechanisms of the PSO-based PNN are explored: the structural optimization is realized via PSO whereas in case of the parametric optimization we proceed with a standard least square method-based learning. To evaluate the performance of the PSO-based PNN, the model is experimented with using Gas furnace process data, and pH neutralization process data. For the characteristic analysis of the given entire data with non-linearity and the construction of efficient model, the given entire system data is partitioned into two type such as Division I(Training dataset and Testing dataset) and Division II(Training dataset, Validation dataset, and Testing dataset). A comparative analysis shows that the proposed PSO-based PNN is model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Journal of the korean society of oceanography
• /
• v.37 no.3
• /
• pp.125-133
• /
• 2002

Water and salt budgets in the Yellow Sea and Bohai are analyzed based on the historical data and CTD data collected recently using box models. The amounts of volume transport and of water exchange across the boundary between the Yellow and East China Seas are estimated to be 2,330-2,840 $\textrm{km}^3$/yr and 109-133 $\textrm{km}^3$/yr, respectively, from the one-layer box model. Corresponding water residence time is 5-6 years. In the Bohai, water residence time is twice as long as that in the Yellow Sea, suggesting that the Yellow Sea and Bohai cannot be considered as a single system in the view of water and salt budgets. The results indicate that water and salt budgets in the Yellow Sea depend almost only on the water exchange between the Yellow and East China Seas. The computation with the coupled two-layer model shows that water residence time is slightly decreased to 4-5 years for the Yellow Sea. In order to reduce uncertainties for the budgeting results the amount of the discharge from the Changjiang that enters into the Yellow Sea, the vertical advection and vertical mixing fluxes across the layer interface have to be quantified. The decreasing trend of the annual Yellow River outflow is likely to result that water residence time is much longer than the current state, especially for the Bohai. The completion of the Three Gorges dam on the Changjiang may be change the water and salt budgets in the Yellow Sea. It is expected that cutting back the discharge from the Changjiang by 10% through the dam would increase water residence time by about 10%.

• Transactions of Materials Processing
• /
• v.21 no.1
• /
• pp.49-57
• /
• 2012

It is difficult to estimate the properties of multilayered sheet because they are composed of one or more different materials. Plastic deformation behavior of the multilayered sheet is quite different as compared to each material individually. The deformation behavior of multilayered sheet should be investigated in order to prevent forming defects and to predict the properties of the formed part. In this study, the mechanical properties and formability of stainless steel-aluminum-magnesium multilayered sheet were investigated. The multilayered sheet needs to be deformed at an elevated temperature because of its poor formability at room temperature. Uniaxial tensile tests were performed at various temperatures and strain rates. Fracture patterns changed mainly at a temperature of $200^{\circ}C$. Uniform and total elongation of multilayered sheet increased to values greater than those of each material when deformed at $250^{\circ}C$. The limiting drawing ratio (LDR) was obtained using a circular cup deep drawing test to measure the formability of the multilayered sheet. A maximum value for the LDR of about 2 was achieved at $250^{\circ}C$, which is the appropriate forming temperature for the Mg alloy. Fracture patterns on a circular cup and thickness of formed part were predicted by a rigid-viscoplastic FEM analysis. Two kinds of modeling techniques were used to simulate deep drawing process of multilayered sheet. A single-layer FE-model, which combines the three different layers into a macroscopic single layer, predicted well the thickness distribution of the drawn cup. In contrast, the location and the time of fracture were estimated better with a multi-layer FE model, which used different material properties for each of the three layers.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.3
• /
• pp.179-184
• /
• 2022

In this paper, we are going to study how we can automatically determine the number of convolutional filters for the optimal model without a search algorithm. This paper proposes HSE Block by connecting SE Block proposed in SENet to a convolutional neural network and connecting a convolutional neural network not learned at the bottom. An experiment was conducted to increase the number of filters by one per 3 epoch using two datasets for the HSEBlock model and to increase the number of filters by the value in the filter. Based on this experiment, the model was constructed with multi-layer HSE Block instead of layer HSE Block, and the experiment was carried out using a dataset that was more difficult to learn than the one used in the previous experiment. The effect of HSE Block was verified by conducting an experiment with the number of HSE Blocks set to 2, 3, 4, and 5 on a dataset that is more difficult to learn than before.