• Polymer(Korea)
• /
• v.26 no.4
• /
• pp.523-534
• /
• 2002

A new hydroxypropylcellulose (HPC) capable of exhibiting reflection colours in the temperature ranges of about 60-$130^{\circ}C$ and acrylic acid esters of HPC (ESs) with degree of esterification (DE) ranging from 1 to 3 were synthesized. The crosslinked ES films with the optical pitch ($\lambda_m$) ranging throughout the visible region were also prepared by exposing thermotropic cholesteric phases of ESs with a DE of more than 2 to UV light at $50^{\circ}C$. The thermal and optical properties for both the uncrosslinked and crosslinked samples and the swelling behavior of the crosslinked films in acetone were investigated. The $\lambda_m$'s of ESs, as well as HPC itself, increased with temperature. However, the $\lambda_m$'s of ESs were larger than of HPC at the same temperature and decreased with increasing DE. The temperature dependence of $\lambda_m$of the crosslinked samples was much weaker than that of ESs. Moreover, in contrast with ESs that exhibit a decrease of the isotropization temperature with increase in the DE, the networks were found to decompose at about $210^{\circ}C$, giving no transition to an isotropic state. The crosslinked samples exhibited an anisotropic swelling, suggesting that the two-dimensional crosslinking preferentially performs between ES molecules.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2011.02a
• /
• pp.444-444
• /
• 2011

Recently, patterned magnetic films and elements attract a wide interest due to their technological potentials in ultrahigh-density magnetic recording and spintronic devices. Among those patterned magnetic structures, magnetic anti-dot patterning induces a strong shape anisotropy in the film, which can control the magnetic properties such as coercivity, permeability, magnetization reversal process, and magneto-resistance. While majority of the previous works have been concentrated on anti-dot arrays with a single magnetic layer, there has been little work on multilayered anti-dot arrays. In this work, we report on study of the magnetic properties of bilayered anti-dot system consisting of upper perforated Co layer of 40 nm and lower continuous Ni layer of 5 nm thick, fabricated by photolithography and wet-etching processes. The magnetic hysteresis (M-H) loops were measured with a superconducting-quantum-interference-device (SQUID) magnetometer (Quantum Design: MPMS). For comparison, investigations on continuous Co thin film and single-layer Co anti-dot arrays were also performed. The magnetic-domain configuration has been measured by using a magnetic force microscope (PSIA: XE-100) equipped with magnetic tips (Nanosensors). An external electromagnet was employed while obtaining the MFM images. The MFM images revealed well-defined periodic domain networks which arise owing to the anisotropies such as magnetic uniaxial anisotropy, configurational anisotropy, etc. The inclusion of holes in a uniform magnetic film and the insertion of a uniform thin Ni layer, drastically affected the coercivity as compared with single Co anti-dot array, without severely affecting the saturation magnetization ($M_s$). The observed changes in the magnetic properties are closely related to the patterning that hinders the domain-wall motion as well as to the magneto-anisotropic bilayer structure.

• The Journal of Engineering Geology
• /
• v.3 no.2
• /
• pp.125-148
• /
• 1993

Groundwater flow in fractured rock masses is controlled by combined effects of fracture networks, state of geostafic stresses and crossflow between fractures and rock matrix. Furthermore the scaie dependent, anisotropic properties of hydraulic parameters results mainly from irregular paftems of fracture system, which can not be evaluated properly with the methods available at present. The basic assumpfion of discrete fracture network model is that groundwater flows only along discrete fractures and the flow paths in rock mass are determined by geometric paftems of interconnected fractures. The characteristics of fracture distribution in space and fracture hydraulic parameters are represented as the probability density functions by stochastic simulation. The discrete fracture network modelling was aftempted to characterize the groundwater flow in the vicinity of existing large cavems located in Wonjeong-ri, Poseung-myon, Pyeungtaek-kun. The fracture data of $1\textrm{km}^2$ area were analysed. The result indicates that the fracture sets evaluated from an equal area projection can be grouped into 6 sets and the fracture sizes are distributed in longnormal. The conductive fracture density of set 1 shows the highest density of 0.37. The groundwater inflow into a carvem was calculated as 29ton/day with the fracture transmissivity of $10^{-8}\textrm{m}^2/s$. When the fracture transmissivity increases in an order, the inflow amount estimated increases dramatically as much as fold, i.e 651 ton/day. One of the great advantages of this model is a forward modelling which can provide a thinking tool for site characterization and allow to handle the quantitative data as well as qualitative data.

• Composites Research
• /
• v.36 no.1
• /
• pp.48-52
• /
• 2023

Fiber-reinforced composites have anisotropic material properties, so the mechanical properties of composite structures can vary depending on the stacking sequence. Therefore, it is essential to design the proper stacking sequence of composite structures according to the functional requirements. However, depending on the manufacturing condition or the shape of the structure, there are many cases where the designed stacking angle is out of range, which can affect structural performance. Accordingly, it is important to analyze the stacking angle in order to confirm that the composite structure is correctly fabricated as designed. In this study, the stacking angle was predicted from real cross-sectional images of fiber-reinforced composites using convolutional neural network (CNN)-based deep learning. Carbon fiber-reinforced composite specimens with several stacking angles were fabricated and their cross-sections were photographed on a micro-scale using an optical microscope. The training was performed for a CNN-based deep learning model using the cross-sectional image data of the composite specimens. As a result, the stacking angle can be predicted from the actual cross-sectional image of the fiber-reinforced composite with high accuracy.