• Journal of Korea Foundry Society
• /
• v.13 no.4
• /
• pp.350-358
• /
• 1993

Aluminum base composites reinforced with various amount of SiC particles and Mg contents have been investigated by different fabrication method for twenty-years. In this paper, how the decomposition and dissolution behaviors of $SiCp(20{\mu}m)$ in the melt of Al composites arised was studied. As the results, the decomposition and dissolution of SiCp into the melt of Al composites increased with increase of the temperature above $720^{\circ}C$, and holding time at a given melting temperature. Because SiC is thermodynamically unstable in this Al-SiCp composite at temperature above the liquidus, SiCp dissolves and reacts with Al in matrix to form $Al_4C_3$ according to following chemical equation $4Al+3SiC{\rightarrow}Al_4C_3+3Si$, Si decomposed and dissolved from SiCp increases Si content of matrix, while liquidus temperature of matrix decrease with increase of SiC content in matrix. The hardness of SiCp decreased with increase of the melting temperature, the hardness of the matrix /particle interface increased with increase of the melting temperature due to increase of the $Mg_2Si$ and $Al_4C_3$ intermetallic compounds, etc.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2002.11a
• /
• pp.91-91
• /
• 2002

• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.125-131
• /
• 2021

In recent years, Convolutional Neural Networks (CNNs) have achieved outstanding performance in the fields of computer vision such as image classification, object detection, visual quality enhancement, etc. However, as huge amount of computation and memory are required in CNN models, there is a limitation in the application of CNN to low-power environments such as mobile or IoT devices. Therefore, the need for neural network compression to reduce the model size while keeping the task performance as much as possible has been emerging. In this paper, we propose a method to compress CNN models by combining matrix decomposition methods of LR (Low-Rank) approximation and CP (Canonical Polyadic) decomposition. Unlike conventional methods that apply one matrix decomposition method to CNN models, we selectively apply two decomposition methods depending on the layer types of CNN to enhance the compression performance. To evaluate the performance of the proposed method, we use the models for image classification such as VGG-16, RestNet50 and MobileNetV2 models. The experimental results show that the proposed method gives improved classification performance at the same range of 1.5 to 12.1 times compression ratio than the existing method that applies only the LR approximation.

• Journal of the Mineralogical Society of Korea
• /
• v.15 no.4
• /
• pp.243-257
• /
• 2002

Ca-type and Na-type bentonites show the great difference of some physicochemical properties. Na exchanged bentonite is mainly used for the foundry and construction materials in domestic utilization. This study tries to identify in detail the differences of some physicochemical properties and thermal properties between Ca-type and Na-type bentonites. Also the adsorption behavior and interlayer expansion for the HDTMA (Hexadecyltrimethylammonium) exchanged and CP (Cetylprydinium) exchanged Ca-type and Na-type bentonites were compared. Na-type bentonite shows the strong alkaline property, high viscosity and swelling compared to Ca-type bentonite. However, two types are very similar for the cation exchange capacity and MB (Methylene Blue) adsorption. The decomposition of adsorbed and interlayer water of Na-type bentonite is caused in the lower temperature than Ca-type bentonite. And Ca-type bentonite shows the decomposition of structural water in the lower temperature than Na-type bentonite. The interlayer expansion of montmorillonite resulted to the intercalation of HDTMA and CP into bentonite is so strongly caused from 12～15 $\AA$ to $40\AA$ (basal spacing). HDTMA-bentonite is almost expanded to $37～38\AA$ when 200% CEC equivalent amount of HDTMA is added, and CP-bentonite is fullly expanded to 40 $\AA$ in the 140% CEC equivalent amount of CP It means that CP causes the stronger interlayer expansion of montmorillonite and easier adsorption than HDTMA. Adsorption behaviors of CP into bentonite is so stable and continuously sorbed in the proportion to the treatment of amount until 200% of the CEC equivalents. CP-bentonite shows the same adsorption behavior regardless of Ca-type or Na-type montmorillonite.

• Membrane and Water Treatment
• /
• v.11 no.3
• /
• pp.195-200
• /
• 2020

In this study, a biocatalyst composite was prepared by immobilizing oxidoreductases onto Cu-activated zeolite to facilitate biochemical decomposition of 4-chlorophenol (4-CP). 4-CP monooxygenase (CphC-I) was cloned from a 4-CP degrading bacterium, Pseudarthrobacter chlorophenolicus A6, and then overexpressed and purified. Type X zeolite was synthesized from non-magnetic coal fly ash using acetic acid treatment, and its surfaces were coated with copper ions via impregnation (Cu-zeolite). Then, the recombinant oxidative and reductive enzymes were immobilized onto Cu-zeolite. The enzymes were effectively immobilized onto the Cu-zeolite (79% of immobilization yield). The retained catalytic activity of CphC-I after immobilization was 0.3423 U/g-Cu-zeolite, which was 63.3% of the value of free enzymes. The results of this study suggest that copper can be used as an effective enzyme immobilization binder because it provides favorable metalhistidine binding between the enzyme and Cu-zeolite.

• Journal of Korean Society of Environmental Engineers
• /
• v.27 no.3
• /
• pp.273-279
• /
• 2005

The degradation of o-chlorophenol(o-CP) in the presence of NaOH and the byproducts formed were investigated in a supercritical water(SCW) destruction process. The conversion of o-CP in the absence of NaOH was less than 20%, however it showed 100% conversion in the presence of NaOH(mole ratio[NaOH]/[o-CP] over than 2) with a residence time of less than 1 second. The formation of PAHs and the phenolic compounds formed were decreased in the presence of NaOH. The results revealed that the formation of byproducts during the destruction of o-CP in SCW was effected by the addition of NaOH. Phenol, cresols, chlorinated phenols, PAHs, p,p'-dihydroxybiphenyl and oxygenated polyaromatic compounds such as 1-indanone, dibenzofuran and dibenzo-p-dioxin were detected in both conditions(presence and absence of NaOH). At the same time, in the presence of NaOH, 2-ethylphenol, o-hydroxyacetophenone, hydroquinone, 4-allylphenol, 3-phenoxyphenol and 4,4'-oxybisphenol were also detected. The observed results suggest that the destruction of o-CP in SCW with NaOH occurs through a number of complicated reaction pathways. Dibenzofuran and dibenzo-p-dioxin were also detected during destruction of o-CP by SCW. The above observation suggests that there may be a common relationship between the thermal incineration process and SCW decomposition process.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.133-135
• /
• 2020

최근 CNN(Convolutional Neural Network)은 영상 분류, 객체 인식 등 다양한 비전 분야에서 우수한 성능을 보여주고 있으나, CNN 모델의 계산량 및 메모리가 매우 커짐에 따라 모바일 또는 IoT(lnternet of Things) 장치와 같은 저전력 환경에 적용되기에는 제한이 따른다. 따라서, CNN 모델의 임무 성능을 유지하연서 네트워크 모델을 압축하는 기법들이 연구되고 있다. 본 논문에서는 행렬 분해 기술인 저계수행렬 근사(Low-rank approximation)와 CP(Canonical Polyadic) 분해 기법을 결합하여 CNN 모델을 압축하는 기법을 제안한다. 제안하는 기법은 계층의 유형에 상관없이 하나의 행렬분해 기법만을 적용하는 기존의 기법과 달리 압축 성능을 높이기 위하여 CNN의 계층 타입에 따라 두 가지 분해 기법을 선택적으로 적용한다. 제안기법의 성능검증을 위하여 영상 분류 CNN 모델인 VGG-16, ResNet50, 그리고 MobileNetV2 모델 압축에 적용하였고, 모델의 계층 유형에 따라 두 가지의 분해 기법을 선택적으로 적용함으로써 저계수행렬 근사 기법만 적용한 경우 보다 1.5~12.1 배의 동일한 압축율에서 분류 성능이 향상됨을 확인하였다.

• Journal of Powder Materials
• /
• v.12 no.4 s.51
• /
• pp.266-272
• /
• 2005

Transition metal ions($Ni^{2+}$, $Cr^{3+}$ and $V^{5+}$) doped $TiO_2$ nanostructured powders were synthesized by mechanical alloying(MA) to shift the adsorption threshold into the visible light region. The synthesized powders were characterized by XRD, SEM, TEM and BET for structural analysis, UV-Vis and photoluminescence spectrum for the optical study. Also, photocatalytic abilities were evaluated by decomposition of 4-chlorophenol(4CP) under ultraviolet and visible light irradiations. Optical studies showed that the absorption wavelength of transition metal ions doped $TiO_2$ powders moved to visible light range, which was believed to be induced by the energy level change due to the doping. Among the prepared $TiO_2$ powders, $NiO^{2+}$ doped $TiO_2$ powders, showed excellent photooxidative ability in 4CP decomposition.

• Journal of Soil and Groundwater Environment
• /
• v.19 no.3
• /
• pp.47-55
• /
• 2014

Arthrobacter chlorophenolicus A6 possesses several monooxygenases (CphC-I, CphC-II, and CphB) that can catalyze the transformation of 4-chlorophenol (4-CP) to hydroxylated intermediates in the initial steps of substrate metabolism. The corresponding genes of the monooxygenases were cloned, and the competent cells were transformed with these recombinant plasmids. Although CphC-II and CphB were expressed as insoluble forms, CphC-I was successfully expressed as a soluble form and isolated by purification. The specific activity of the purified CphC-I was analyzed by using 4-CP, 4-chlorocatechol (4-CC), and catechol (CAT) as substrates. The specific activities for 4-CP, 4-CC, and CAT were determined to be 0.312 U/mg, 0.462 U/mg, 0.246 U/mg, respectively. The results of this study indicated that CphC-I is able to catalyze the degradation of 4-CC and CAT in addition to 4-CP, which is a primary substrate. This research is expected to provide the fundamental information for the development of an eco-friendly biochemical degradation of aromatic hydrocarbons.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.12C
• /
• pp.1257-1264
• /
• 2006

The carrier frequency offset causes loss of orthogonality between sub-carriers, thus leads to inter-carrier interference (ICI) in the OFDM symbol. This ICI causes severe degradation of the BER performance of the OFDM receiver. In this paper, we propose a new ICI cancellation algorithm which estimates frequency offset at the time-domain by using CP-ICA method to the received sub-carriers phase rotation. This algorithm is based on a statistical blind estimation method, which mainly utilizes the EVD, rotating phase and the $4^{th}-cumulants$. Since our scheme does not need any training and pilot symbol in estimation, we can expect enhanced bandwidth efficiency in OFDM systems. Simulation results show that the proposed frequency offset estimator is more accurate than the other estimators in $0.0<\varepsilon<1.0$.