• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.23 no.11
• /
• pp.1418-1425
• /
• 1999

Flow over a protruding surface is investigated using numerical simulation. We consider flow between two parallel plates with a cube mounted on one side of the channel. As the flow approaches the cube, the adverse pressure gradient produces three-dimensional boundary-layer separation, resulting In the formation of horseshoe vortices. The objective of our study is to clarify both the steady and the unsteady characteristics of the vortex system. As the Reynolds number increases, the structure of the vortices near the cube becomes complex and the number of vortices increases. The distribution of skin friction on the cube-mounted wall reflects the effect of the horseshoe vortices. All these results are consistent with the experimental findings currently available.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.21 no.48
• /
• pp.123-132
• /
• 1998

The group formation problem of the machine and part is a very important issue in the planning stage of cellular manufacturing systems. This paper investigates Self-Organizing Map(SOM) neural networks approach to machine-part grouping problem. We present a two-phase algorithm based on SOM for grouping parts and machines. SOM can learn from complex, multi-dimensional data and transform them into visually decipherable clusters. Output layer in SOM network is one-dimensional structure and the number of output node has been increased sufficiently to spread out the input vectors in the order of similarity. The proposed algorithm performs remarkably well in comparison with many other algorithms for the well-known problems shown in previous papers.

• Journal of Institute of Control, Robotics and Systems
• /
• v.10 no.11
• /
• pp.1006-1011
• /
• 2004

Carbon nanotubes have outstanding properties which make them useful for a number of high-technology applications. Especially, single-walled carbon nanotube (SWNT), working under physical conditions (in aqueous solution) and converting electrical energy into mechanical energy directly, can be a good substitute for artificial muscle. The carbon nanotube structure simulated in this paper is an isotropic cantilever type with an adhesive tape which is sandwiched between two SWNTs. For predicting the geometrical and physical parameters such as deflection, slope, bending moment and induced force with various applied voltages, the analytical model for a 3 layer bimorph nanotube actuator is developed by applying Euler-Bernoulli beam theory. The governing equation and boundary conditions are derived from energy Principles. Also, the brief history of carbon nonotube is overviewed and its properties are compared with other functional materials. Moreover, an electro-mechanical coupling coefficient of the carbon nanotube actuator is discussed to identify the electro-mechanical energy efficiency.

• Journal of Magnetics
• /
• v.5 no.2
• /
• pp.27-34
• /
• 2000

The structure of a series of Cu/Ni is characterized by using a program, SUPREX, to model the x-ray diffraction patterns, multilayers. The samples had nominal layer thickness of 3/3, 7/7, 13.5/13.5, 20/20, 30/30, 50/50, 80/80, 100/100, and 200/200 Angstroms. The diffraction patterns were taken around the (111) peak for the two constituent materials. A kinematical model is used to characterize the diffraction patterns and the parameters for the model are described. An initial model is calculated using initial guesses for the parameters. The model is then fit to the data by reducing $x^2$using the Levenberg-Marquardt algorithm. The samples are shown to be high quality supperlattices.

• Proceedings of the KSME Conference
• /
• 2003.04a
• /
• pp.1043-1047
• /
• 2003

With the advancement of micro-systems and nanotechnology, the need for ultra-precision fabrication techniques has been steadily increasing. In this paper, a novel nano-structure fabrication process that is based on the fundamental understanding of nano-scale tribological interaction is introduced. The process, which is called Mechano-Chemical Scanning Probe Lithography (MC-SPL), has two steps, namely, mechanical scribing for the removal of a resist layer and selective chemical etching on the scribed regions. Organic monolayers are used as a resist material, since it is essential for the resist to be as thin as possible in order to fabricate more precise patterns and surface structures. The results show that high resolution patterns with sub-micrometer scale width can be fabricated on both silicon and various metal surfaces by using this technique.

• Proceedings of the Korean Institute of Surface Engineering Conference
• /
• 2014.11a
• /
• pp.241-241
• /
• 2014

$SnO_2-core/Zn-shell$ heteronanowires were fabricated by a two-step process: thermal evaporation of Sn powders and employing a sputtering technique with a Zn target. X-ray diffraction, high-resolution transmission electron microscopy, and EDX spectra coincidentally indicated that the shell layer comprised the Zn phase. From Gaussian deconvolution studies, we observed that photoluminescence (PL) spectra consisted of yellow, green, and ultraviolet (UV) emission bands, regardless of shell-coating. We speculated the possible mechanisms of these emission peaks.

• Journal of Integrative Natural Science
• /
• v.2 no.3
• /
• pp.211-214
• /
• 2009

A new porous silicon (PSi) microcavity sensor for the detection of volatile organic compounds (VOCs) was developed. PSi microcavity sensor exhibiting unique reflectivity was successfully obtained by an electrochemical etching of silicon wafer. When PSi was fabricated into a structure consisting of two high reflectivity muktilayer mirrors separated by an active layer, a microcavity was formed. This PSi microcavity is very sensitive structures. Reflection spectrum of PSi microcavity indicated that the full-width at half-maximum (FWHM) was of 10 nm and much narrower than that of fluorescent organic molecules or quantum dot. The detection of volatile organic compounds (VOCs) using PSi microcavity was achieved. When the vapor of VOCs condensed in the nanopores, the refractive indices of entire particle increased. When PSi microcavity was exposed to acetone, ether, and toluene, PSi microcavity in reflectivity was red shifted by 28 nm, 33 nm, and 20 nm for 2 sec, respectively.

• Proceedings of the KSME Conference
• /
• 2004.04a
• /
• pp.410-414
• /
• 2004

Under longitudinal loading continuous fiber reinforced metal matrix composite(MMC) have interpreted an outstanding performance. However, the applicability of continuous fiber reinforced MMCs is somewhat limited due to their relatively poor transverse properties. Therefore, the transverse properties of MMCs are significantly influenced by the properties of the fiber/matrix interface. In this study, elastic-plastic behavior of transversely loaded unidirectional fiber reinforced metal matrix composites investigated by using elastic-plastic finite element analysis. Different fiber placement(square and hexagon) and fiber volume fractions were studied numerically. The interface was treated as three thin layer (with different properties) with a finite thickness between the fiber and the matrix. The analyses were based on a two-dimensional generalized plane strain model of a cross-section of an unidirectional composite by the ANSYS finite element analysis code.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.6
• /
• pp.767-772
• /
• 2018

Recently, some meta-heuristic algorithms, such as GA(Genetic Algorithm) and GP(Genetic Programming), have been used to optimize CNN(Convolutional Neural Network). The CNN, which is one of the deep learning models, has seen much success in a variety of computer vision tasks. However, designing CNN architectures still requires expert knowledge and a lot of trial and error. In this paper, the recent attempts to automatically construct CNN architectures are investigated and analyzed. First, two GA based methods are summarized. One is the optimization of CNN structures with the number and size of filters, connection between consecutive layers, and activation functions of each layer. The other is an new encoding method to represent complex convolutional layers in a fixed-length binary string, Second, CGP(Cartesian Genetic Programming) based method is surveyed for CNN structure optimization with highly functional modules, such as convolutional blocks and tensor concatenation, as the node functions in CGP. The comparison for three approaches is analysed and the outlook for the potential next steps is suggested.

• Journal of Information Processing Systems
• /
• v.6 no.1
• /
• pp.33-42
• /
• 2010

GIS can only be applied to certain areas by storing format. It is subordinate to a system when displaying geographic information data. It is therefore inevitable for GIS to use GML that supports efficient usage of various geographic information data and interoperability for integration and sharing. The paper constructs VisualGML that translates currently-used geographic information such as DXF (Drawing Exchange Format), DWG (DraWinG), or SHP (Shapefile) into GML format for visualization. VisualGML constructs an integrated map pre-process module, which filters geographic information data according to its tag and properties, to provide the flexibility of a mobile device. VisualGML also provides two major GIS services for the user and administrator. It can enable visualizing location search. This is applied with a 3-Layer POI structure for the user. It has trace monitoring visualization through moving information of mobile devices for the administrator.