• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.06a
• /
• pp.1074-1078
• /
• 2005

In precision positioning applications, such as scanning tunneling microscopy and diamond turning machines [1], it is often required that actuators have nanometer resolution in displacement, high stiffness, and fast frequency response. These requirements are met by the use of piezoceramic actuators. A major limitation of piezoceramic actuators, however, is their lack of accuracy due to hysteresis nonlinearity and drift. The maximum error due to hysteresis can be as much as 10-15% of the path covered if the actuators are run in an open-loop fashion. Hence, the accurate control of piezoceramic actuators requires a control strategy that incorporates some form of compensation for the hysteresis. One approach is to develop an accurate model of the hysteresis and the use the inverse as a compensator. The Preisach model has frequently been employed as a nonlinear model for representing the hysteresis, because it encompasses the basic features of the hysteresis phenomena in a conceptually simple and mathematically elegant way. In this paper, a new numerical inversion scheme of the Preisach model is developed with an aim of compensating hysteresis in piezoceramic actuators. The inversion scheme is implemented using the first-order reversal functions and is presented in a recursive form. The inverted model is then incorporated in an open-loop control strategy that regulates the piezoceramic actuator and compensates for hysteretic effects. Experimental results demonstrate satisfactory regulation of the position of the piezoceramic actuator to the desired trajectories.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.19 no.1
• /
• pp.122-131
• /
• 1995

The research trends for metal matrix composites have been on basic mechanical properties, fatigue behavior after aging and fractographic observations. In this study, the fatigue crack initiation as well as the fatigue crack growth behavior and the fracture mechanism were investigated through observations of the fracture surface on silicon carbide particles reinforced aluminum metal matrix composites(SiCp/Al). Based on the fractographic study done by scanning electron microscope and replica, crack growth path model and fracture mechanism are presented. The mechanical properties, such as the tensile strength, yield strength and elongation of SiCp/Al composites are improved in a longitudinal direction, however, the fatigue life is shorter than the basic Al6061 alloys. From fractographic observations, it is found that the failure mode is ductile in basic Ai6061 alloys. And because some SiC particles were pulled out from the matrix and a few SiC particles could be seen on the fracture surface of SiCp/Al, crack growth paths are believed to follow the interface of the matrix and its particles.

• Smart Structures and Systems
• /
• v.6 no.2
• /
• pp.91-100
• /
• 2010

An experimental study was carried out to investigate the shape control of plates via embedded shape memory alloy (SMA) wires. An extensive body of literature proposes the use of SMA wires to actively modify the shape or stiffness of a structure; in most cases, however, the study focuses on modeling and little experimental data is available. In this work, a simple proof of concept specimen was built by attaching four prestrained SMA wires to one side of a carbon fiber laminate plate strip. The specimen was clamped at one end and tested in an environmental chamber, measuring the tip displacement and the SMA temperature. At heating, actuation of the SMA wires bends the plate; at cooling deformation is partially recovered. The specimen was actuated a few times between two fixed temperatures $T_c$ and $T_h$, whereas in the last actuation a temperature $T_f$ > $T_h$ was reached. Contrary to most model predictions, in the first actuation the transformation temperatures are significantly higher than in the following cycles, which are stable. Moreover, if the temperature $T_h$ is exceeded, two separate actuations occur during heating: the first follows the path of the stable cycles; the second, starting at $T_h$, is similar to the first cycle. An interpretation of the phenomenon is given using some differential scanning calorimeter (DSC) measurements. The observed behavior emphasizes the need to build a more comprehensive constitutive model able to include these effects.

• Journal of Korea Foundry Society
• /
• v.18 no.6
• /
• pp.550-554
• /
• 1998

The microstructures of rapidly solidified binary Ti-Al alloys containing $45{\sim}58\;at%Al$ have been studied using C/S (carbon/sulfur), N/O (nitrogen/oxygen) analyser, X-ray fluorescence spectrometer (XRF), X-ray diffractometer (XRD), optical microscope (OM) and scanning electron microscope (SEM). The phases present in the alloys and their distribution were found to be a sensitive function of Al content. Essentially single-phase (${\gamma}$) microstructures were observed to alloys with 45 at%Al, 55 at%Al and 58 at%Al. In other content alloys, two phase (${\alpha}_2$, ${\gamma}$) microstructures were observed. The 48 at%Al, 52 at%Al alloys contain (${\gamma}+{\alpha}_2$) phase and ${\alpha}_2$ phase. These results indicate that rapid solidification affect the solidification path, then metastable phase forming during solidification.

• Journal of Welding and Joining
• /
• v.21 no.2
• /
• pp.102-110
• /
• 2003

A visual sensor consisted of polygonal mirror, laser, and CCD camera was proposed to measure the distance to the weld joint for recognizing the joint shape. To scan the laser beam of the sensor onto an object, 8-facet polygonal mirror was used as the rotating mirror. By locating the laser and the camera at axi-symmetrical positions around the mirror, the synchronized-scan condition could be satisfied even when the mirror was set to rotate through one direction continuously, which could remove the inertia effect of the conventional oscillating-mirror methods. The mathematical modelling of the proposed sensor with the optical triangulation method made it possible to derive the relation between the position of an image on the camera and the one of a laser light on the object. Through the geometrical simulation of the proposed sensor with the principal of reflection and virtual image, the optical path of a laser light could be predicted. The position and direction of the CCD camera were determined based on the Scheimpflug's condition to fit the focus of any image reflected from an object within the field of view. The results of modelling and simulation revealed that the proposed visual sensor could be used to recognize the weld joint and its vicinity located within the range of the field of view and the resolution. (Received February 19, 2003)

• Applied Science and Convergence Technology
• /
• v.23 no.6
• /
• pp.317-327
• /
• 2014

In order to develop nano-devices with much lower power consumption for beyond-CMOS applications, the fundamental understanding and precise control of the electronic properties of ultrathin transition metal oxide (TMO) films are strongly required. The metal-insulator transition (MIT) is not only an important issue in solid state physics, but also a useful phenomenon for device applications like switching or memory devices. For potential use in such application, the electronic structures of MIT, observed for TMO nano-structures, have been investigated using a synchrotron radiation angle-resolved photoelectron spectroscopy system combined with a laser molecular beam epitaxy chamber and a scanning photoelectron microscopy system with 70 nm spatial resolution. In this review article, electronic structures revealed by soft X-ray nano-spectroscopy are presented for i) polarity-dependent MIT and thickness-dependent MIT of TMO ultrathin films of $LaAlO_3/SrTiO_3$ and $SrVO_3/SrTiO_3$, respectively, and ii) electric field-induced MIT of TMO nano-structures showing resistance switching behaviors due to interfacial redox reactions and/or filamentary path formation. These electronic structures have been successfully correlated with the electrical properties of nano-structured films and nano-devices.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2014.02a
• /
• pp.248.1-248.1
• /
• 2014

We present results from an experimental study of conductivity change of poly (3,4-ethylenedioxythiophene) poly(styrenesulfonate) (PEDOT:PSS) film according to the surface structuring. We demonstrate that the patterned structure was enhanced with approximately five times conductivity in comparison with non structure of PEDOT:PSS film. In order to patterning, we have fabricated polystyrene (PS) colloidal monolayer as a template with sphere diameter of 780nm and 1.8um. Structure has honeycomb shape and it provide shorter path way to flowing of electron. Pattern size was controlled by PS diameter and varied by Transformer Coupled Plasma (TCP) etching system. Conductivity was converted from sheet resistance which measured by 4-point prove. Film thickness was derived using Field Emission Scanning Electron Microscopy (FE-SEM) images.

• Journal of the Korean Society for Precision Engineering
• /
• v.27 no.12
• /
• pp.99-106
• /
• 2010

Gas turbines operation for power generation increased rapidly since 1990 due to the high efficiency in combined cycle, relatively low construction cost and low emission. But the operation and maintenance cost for gas turbine is high because the expensive superalloy hot gas path parts should be repaired and replaced periodically This study analyzed the initiation and propagation of the crack at the gas turbine blades which are coated with MCrAIY as a bond coat and TBC as a top coat. The sample blades had been serviced at the actual gas turbines for power generation. Total 7 sets of blades were analyzed and they have different EOH(equivalent operation hour). Blades were sectioned and the cracking distribution were measured and analyzed utilizing SEM(scanning electron microscope) and optical microscope. The blades which had 52,000 EOH of operation had cracks at the substrate and the maximum depth was 0.2 mm. Most of the cracks initiated at the boundary layer between TBC and bond coat and propagated down to the bond coat. Once bond coat is cracked, the base metal is exposed to the oxidation condition and undergoes notch effect. Under this environment, the crack branched at the inter-diffusion layer and propagated to the substrate. Critical cracks affecting the blade life were analyzed as those on suction side and platform.

• 한국신재생에너지학회:학술대회논문집
• /
• 2009.11a
• /
• pp.132-132
• /
• 2009

To achieve the commercialization of PEM fuel cell, the durability problem must be solved. Recently, many researchers have focused on this durability problem and degradation studies about membrane and electrode have been reported. But durability characteristics of gas diffusion layer is not much reported yet. Durability of GDL is very important to maintain the performance of PEM fuel cell because the main function of GDL is a path of fuel and water and the GDL degradation causes the loss of the GDL function. In this study, the degradation of GDL, especially, the mechanical degradation process was investigated with the leaching test. The effect of water dissolution was observed through the test and the amount of GDL degradation was measured with various measurement methods such as weight measurement, static contact angle measurement, scanning electron microscope. After 2,000 hours test, the GDL showed structural damage and loss of hydrophobicity.

• Proceeding of EDISON Challenge
• /
• 2014.03a
• /
• pp.1-13
• /
• 2014

Potential Energy Surface(PES)를 양자 계산을 통해 알아내는 것은 화학 반응을 이해하는 데에 큰 도움이 된다. 이를테면 Transition State(TS)의 configuration을 알 수 있고, 따라서 reaction path와 활성화 에너지 값을 예측하여, 진행시키고자 하는 화학반응의 이해를 도울 수 있다. 하지만 PES를 그리기 위해서는 해당 분자의 다양한 configuration에 대한 singlet point energy 계산이 필요하기 때문에, 계산적인 측면에서 많은 비용을 요구한다. 따라서 product와 reactant의 구조와 같은 critical point의 정보를 이용하여 최소한의 configuration을 sampling하여 전체 PES를 재구성하는 기계학습 알고리즘을 개발하여 다차원 PES 상에서의 화학반응의 예측을 가능하게 하고자 한다. 본 연구에서는 Barbaralane의 두 안정화 된 구조의 critical point로 하여 이 주변을 random normal distribution하여, B3LYP/6-31G(d) level의 DFT 계산을 통해 relaxed scanning하여 구조와 에너지를 구하였으며, 이 정보를 Support Vector Regression(SVR) 알고리즘을 적용하여 PES를 재구현하였으며, 반응경로와 TS의 구조 그리고 활성화 에너지를 구하였다. 또한 본 기계학습 알고리즘을 바닥상태에서 일어나는 반응이 아닌, 들뜬 상태와 전자 구조가 변하는 화학반응, avoid crossing, conical intersection과 같은 Non-adiabatic frame에서 일어나는 현상에 적용 가능성을 논하고자 한다.