Transparent copolyester/organoclay nanocomposites prepared by in situ intercalation polymerization: Synthesis,characterization, and properties

In this study, amorphous poly(ethylene terephthalate‐co‐1,3/1,4‐cyclohexylenedimethylene terephthalate) (PETG)/organoclay nanocomposites was synthesized by the in situ intercalation polymerization of terephthalic acid, ethylene glycol, 1,3/1,4‐cyclohexanedimethanol, and organoclay. The organoclay was obtained by modifying sodium montmorillonite (clay) with hexadecyl triphenylphosphonium bromide. The thermal, mechanical, optical, and gas barrier properties of these PETG nanocomposites with various organoclay contents (0–3 wt%) were discussed. The differential scanning calorimetry and X‐ray analyses revealed that all of the nanocomposites were amorphous. X‐ray diffraction and transmission electron micrographs showed that the organoclay was well dispersed in the polymer matrix, although some parts of the agglomerated layers remained on the scale of several hundreds of nanometers. The thermal stability and the mechanical property of the nanocomposites increased with organoclay content. The optical transmittances of nanocomposites that contained 0.5, 1, and 3 wt% of organoclay were 86.8%, 84.4%, and 77.4%, respectively. The oxygen transmission rate of the nanocomposite that contained 3 wt% of organoclay was about 50% of the PETG base polymer. POLYM. COMPOS., 2011. © 2010 Society of Plastics Engineers     

The development of automated solder bump inspection using machine vision techniques

Visual inspection is an important task in the manufacturing processes for integrated circuit boards. In this paper, we focus on the solder bump inspection problem, and an automated visual inspection method using machine vision techniques is proposed. The solder bump inspection method consists of image grabbing, image preprocessing, feature extraction, and defect detection and classification. Five defect types of solder bumps to be inspected are bridging solder, excess solder, incomplete solder, non-wetting, and missing solder. The solder area, the number of edge pixels, the deviation from center, and the deformation ratio are used as the features for solder bump defect detection and classification. The proposed method is a hybrid algorithm, and it consists of two stages: the training stage and the inspection stage. The experimental results show that the proposed method is effective in detecting defects of solder bumps.     

Fire performance and post‐fire mechanical properties of polymer composites coated with hybrid carbon nanofiber paper

In this study, glass fiber reinforced polyester composites were coated with carbon nanofiber/clay/ammonium polyphosphate (CCA) paper and carbon nanofiber/exfoliated graphite nanoplatelets/ammonium polyphosphate (CXA) paper. The composites were exposed to a heat flux of 35 kW/m² during the cone calorimeter testing. The testing results showed a significant reduction in both heat release rates and mass loss rates. The peak heat release rate (PHRR) of CCA and CXA composite samples in the major decomposition period are 23 and 34% lower than the control sample, respectively. The time to reach the PHRR for the CCA and CXA composite samples are ～ 125% longer than the control sample. After the composite samples were exposed to heat for different time periods, their post‐fire mechanical properties were determined by three‐point bending testing. The three‐point bending testing results show that the composite samples coated with such hybrid papers exhibit more than 20% improvement in mechanical resistance at early stages of combustion. The mechanism of hybrid carbon nanofiber paper protecting the underlying laminated composites is discussed. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

An active device based wide tuning range CMOS transconductor

A CMOS transconductor for wide tuning range filter application is presented. The linear transconductor is designed based on the flipped-voltage follower (FVF) circuit and can work in the weak, moderate, and strong inversion regions to maximize the transconductance tuning range. The transconductance tuning can be achieved by changing the bias current of the active resistor, and a ratio of 28 is obtained. The transconductor was evaluated by using TSMC 0.18 μm CMOS process, and the total harmonic distortion (THD) of −56 dB can be obtained by giving a 12 MHz 0.4 Vpp input swing signal. In the design, the maximum power consumption is 2 mW with the transconductance of 1.1 mS under a 1.8 V supply voltage.     

Critical design issues for high-power GaN/AlGaN anti-serial Schottky varactor frequency triplers

In this paper the importance of a new design variable for high power anti-serial Schottky varactors, the aluminum composition of the AlGaN barrier layer, is demonstrated. AlGaN/GaN varactors containing either (1) a high-doped/low-doped GaN region or (2) just a low doped GaN region have been compared demonstrating that the selection of the device structure also depends on the amplitude of the input signal being tripled in frequency. Stronger susceptance modulation is exhibited in AlGaN/GaN ASVs made from Ga-face polar material compared to N-face polar material. Results indicate choosing the proper aluminum composition results in 27% conversion efficiency with an input signal of 5 GHz and over 7% conversion efficiency with an input signal of 60 GHz along with optimization trends. With input voltage amplitudes over 10 V an AlGaN/GaN structure with 15% Al provides greater conversion efficiency than one with 5% Al. Power absorbed in the varactor also increases as aluminum percent increases affecting reliability and power transfer. Results of a GaN ASV performing as a frequency tripler for fundamental frequencies up to 110 GHz indicate an advantage to using an AlGaN/GaN epi-structure over only a GaN epi-structure.     

Estimation of the Mechanical Property of CNT Ropes Using Atomistic-Continuum Mechanics and the Equivalent Methods

The development in the field of nanotechnology has prompted numerous researchers to develop various simulation methods for determining the material properties of nanoscale structures. However, these methods are restricted by the speed limitation of the central processing unit (CPU), which cannot estimate larger-scale nanoscale models within an acceptable time. Thus, decreasing the CPU processing time and retaining the estimation accuracy of physical properties of nanoscale structures have become critical issues. Accordingly, this study aims to decrease the CPU processing time and complexity of large nanoscale models by utilizing, atomistic-continuum mechanics (ACM) to build an equivalent model of carbon nanotubes (CNTs). The results of tensile and modal analyses agree with previous experimental results indicating that the ACM model can accurately describe mechanical properties. This study also adopted three definitions of cross-sectional area to explore whether the structure properties of CNT ropes depends on the definitions of cross-sectional area. Results indicate that the Young’s modulus distribution based on the circumcircle assumptions agrees well with most of the experimental results. Hence, most experimental methods adopted the circumcircle to obtain the Young’s modulus of the CNT ropes. The circumcircle assumption involves the distribution of the tubes and the gap between each tube. The ratio between the gap and tube areas becomes a stable value when the diameter of the CNT ropes is increased. Therefore, a larger diameter of CNT ropes that represents the Young’s modulus becomes a stable value, as mentioned in literature. This study also investigated the equivalent solid, shell, and beam models to generate similar mechanical behaviors with the ACM model. The similar mechanical behavior of the equivalent model includes the model under tensile, torsion, or shear external loading. These equivalent models can significantly reduce the required total element number and CPU processing time to investigate a larger nanoscale structure.     

基于BP神经网络的手绘电路图形符号识别技术

首先介绍了手绘电路图形符号扫描输入和图像信息去噪压缩处理方法,并离散化成标准的二维矩阵形式.然后引入一个三层前馈型BP神经网络,采用Delta迭代训练算法,学习图像矩阵输入与元件符号输出之间的映射关系,并着重阐述了神经网络学习的过拟合问题.最后通过MATLAB编程仿真和交叉验证表明,提出的BP神经网络结构及其训练算法合理、有效和实用.