A new way of strengthening and toughening for carbon fiber reinforced polyphenylene sulfide (CF/PPS) composites via matrix modification

The effect of pressure-induced flow(PIF) processing on the mechanical properties of noncontinuous carbon fiber(CF) reinforced polyphenylene sulfide(PPS) composites was investigated. A series of CF/PPS composites under different processing conditions were prepared through PIF-processing. SEM observations showed that the interfaces adhesion between CFs and PPS became stronger and ductile fracture mainly occurred in PPS matrix. This brought to a great increase of both strength and toughness by about 2 folds, when the composites were processed at 240 ℃ and under 263 MPa. The results in differential scanning calorimetry(DSC) and X-ray diffraction(XRD) measurements indicated more regular crystalline structures and orientation of lamellae formed during PIF-processing.     

Automatic Quality Inspection of Bakery Products Based on Shape and Color Information

基于形状和颜色信息的烘焙产品的自动质量检测

Babar Khan,韩芳*,王直杰,Ather Iqbal

（东华大学 信息科学与技术学院,上海 201620）

创新点说明：

利用支持向量机和基于生物启发式HMAX模型的形状辨识器与RGB对抗颜色通道辨识器集成而形成一种有效的智能分类系统,并将其用于烘焙产品的质量检测。

研究目的：

本文要对烘焙产品进行更有效的质量检测,需要提出一种智能方法。

研究方法：

本文采用了支持向量机、基于生物启发式HMAX模型的形状辨识器与RGB对抗颜色通道辨识器集成,形成一种高级自动分类系统,对烘焙产品根据其形状和颜色（烘焙时间）进行对应分类。

结果：

利用本文方法对烘焙产品进行质量检测的结果与其它方法相比较,发现本文提出的基于生物启发的机器视觉系统运行更加精确和有效,同时发现采用One Versus One支持向量机和Directed Acyclic Graph支持向量机可以得到最大的分类精确率。本文方法对饼干形状和颜色的辨识分别达到了95%和100%的分类精确率。同时,算法稳定。

结论：

本文方法因其更加可靠和精确可用于代替现有的烘焙产品检测方法。

关键词：烘焙产品质量检测,计算机视觉,HMAX, 对抗颜色通道,RGB颜色辨识器,支持向量机

     

基于用户分层的手持工具感知差异的研究

为探索设计者、供应商和终端用户对手持工具的感知差异,以钳子为研究对象,进行视觉感官试验、触觉感官试验和操作体验测试,运用因子分析法和方差分析法进行对比分析,得出用于钳子评价的语义因子,发现设计者和终端用户对钳子的可用性和便携性方面的感知差异较显著,在使用前后的复杂性和弹性的评价差异较大。     

Synthesis and characterization of poly(hexamethylene terephthalate/hexamethylene oxamide) alternating copolyamide (alt-PA6T/62)

As a representative polyoxamide, poly(hexamethylene oxamide) (PA62) has good comprehensive performance. However, the high T_m (330°C) creates an obstacle for processing. To improve the processability of PA62, poly(hexamethylene terephthalate/hexamethylene oxamide) alternating copolyamide (alt-PA6T/62) was synthesized by hexamethylene diamine-terminated 6T6-diamine and dibutyl oxalate via solution/solid state polycondensation. Random copolyamide (ran-PA6T/62) was also synthesized for comparison. The structure and properties of the copolymer were analyzed by Fourier transform infrared spectroscopy (FT-IR), nuclear magnetic resonance (NMR) spectroscopy, differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), Wide-angle X-ray diffraction (WAXD) and the saturated water absorption test. The NMR results confirm the alternating structure of alt-PA6T/62. The DSC and TGA results demonstrate that the novel alternating copolyamide alt-PA6T/62 (T_m = 321°C, T₅ = 420°C) exhibited better thermal properties than those of ran-PA6T/62 (T_m = 294°C, T₅ = 412°C). The saturated water absorption of alt-PA6T/62 was found to be 3.2 wt%. These results revealed that the novel alt-PA6T/62 had an alternating sequence distribution, showed a high melting point as well as good processability and thermal stability, and possessed low saturated water absorption and excellent dimensional stability.     

Preparation and characterization of high strength and high modulus PVA fiber via dry-wet spinning with cross-linking of boric acid

To improve the mechanical properties of polyvinyl alcohol (PVA) fibers, a series of PVA fibers were prepared via dry-wet spinning with cross-linking of boric acid (BA) (PVA/BA fibers), and using the mixed solvent of dimethyl sulfoxide and water. Moreover, the final PVA/BA fibers were characterized by Fourier transform infrared spectra (FTIR), scanning electron microscopy (SEM), differential scanning calorimetery (DSC), thermogravimetric analyzer (TGA), powder X-ray diffraction (XRD) and yarn strength tester. Furthermore, with the increasing of BA content, FTIR analysis showed that the degree of crosslinking of BA with PVA increased. SEM images of final PVA/BA fibers presented smooth surfaces, and the diameters decreased firstly and then increased. DSC, TGA, and XRD analysis indicated that the melting temperatures, thermal properties and crystallinities first increased and then decreased with the increasing of BA content. In addition, mechanical properties measurements illustrated that the cross-linking existed at an optimal BA content of 0.3 wt%, and PVA/BA-0.3 fiber had the highest tensile strength and Young's modulus of 13.1 ± 0.4 and 360.2 ± 10.4 cN/dtex, respectively.     

Enhanced wear resistance of ultra-high molecular weight polyethylene fibers by modified-graphite oxide

A simple and feasible method to enhance the wear resistance of ultra-high molecular weight polyethylene (UHMWPE) fibers was reported. The graphite oxide (GO) prepared using improved Hummer's method was surface modified with hexadecylamine to improve its compatibility with UHMWPE. Combined with well-dispersion of modified-GO (m-GO) in dichloromethane and the fact that the viscosity of UHMWPE suspension can be decreased by dichloromethane, the well dispersed m-GO/dichloromethane was added into UHMWPE suspension to improve m-GO dispersion in UHMWPE fibers. Finally, UHMWPE fibers with different m-GO concentration were prepared using gel spinning technology. The effect of m-GO concentration on the structure and properties of modified UHMWPE fibers were investigated. The results indicated that the melting temperature and crystallinity of m-GO modified UHMWPE fibers increased with increasing of m-GO concentration, while the fiber's crystal sizes and orientation increased, thus the tensile strength of m-GO modified UHMWPE fibers remained almost undamaged. The introduction of m-GO is beneficial to the formation of smooth transfer film on fiber's surface, which enhanced the self-lubrication of UHMWPE fibers. Compared with pure UHMWPE fiber, the UHMWPE fiber containing 1.5 wt% m-GO had enhanced wear resistance by 55.4% and still maintained high tensile strength of 29.98 cN dtex⁻¹.     

Sequential ion-electron irradiation of zirconium carbide ceramics: Microstructural analysis

Zirconium Carbide (ZrC_x) was irradiated with 10 MeV Au³⁺ ions to a dose of 10 displacements per atoms (dpa) and subsequently with 100 and 300 keV electrons in a transmission electron microscope (TEM). After ion irradiation, dislocation loops were observed in the microstructure and an increase in the number of carbon vacancies was revealed by Raman spectroscopy. Grazing incidence X-ray diffraction (GIXRD) analysis showed that neither amorphization nor oxidation occurred during ion irradiation of the specimen. Subsequent electron irradiation of the pre-implanted ZrC_x foil led to formation of nanosized tetragonal ZrO₂ precipitates (5−10 nm diameter) on the surface of the TEM lamella. The formation of the new oxide phase was not related to the electron beam-induced heating of the specimen, but to electron stimulated oxidation caused by the residual oxygen inside the transmission electron microscope. Changes in size and density of ZrO₂ crystallites were observed between the pristine and ion irradiated ZrC_x regions following electron irradiation, suggesting that the initial microstructure of the ZrC_x substrate played a key role in the nucleation and growth of the oxide islands. The obtained results provide insights into the microstructural response of ZrC_x to different types of radiation and the inadvertent effects of the electron beam during TEM analysis of in-situ and ex-situ ion irradiated ZrC_x. Additionally, the findings of this work suggest a method to prepare local ZrO₂ nanoprecipitates within ZrC_x grains by selective electron beam irradiation.     

同步旋转坐标系下的BLDCM滑模观测器算法设计

为了实现无位置传感器无刷直流电机(BLDCM)矢量控制系统中电机转子位置的准确估计,提出了一种基于同步旋转坐标系的滑模观测器算法。该方法直接在同步旋转坐标系中设计滑模观测器,以获取电机反电动势信息,再通过锁相环技术从估计的反电动势中提取电机转子的速度和位置角度信息。针对滑模观测器的高频抖振问题,采用饱和函数代替滑模观测器的符号函数。最后,通过仿真将所提算法与传统滑模观测器算法对比,并对所提算法进行实验验证。仿真与实验结果表明该算法能够准确跟踪转子的速度和位置,验证了所提算法的正确性与可行性。     

一种改进的BP神经网络在故障诊断中的应用研究

人工神经网络作为一个具有高度非线性映射能力的计算模型,在工程中具有广泛的应用前景.它不需要预选确定样本的数学模型,仅通过学习样本数据即可以进行故障诊断.本文结合计算机综合业务的一些实际问题,探讨了利用改进的BP神经网络进行故障诊断的方法和应该注意的问题,在分析神经网络的基础上提出了基于改进的BP模型神经网络的故障诊断推理方法.结果表明,基于BP神经网络的故障诊断方法是行之有效的.     

球对称n维反应扩散方程的边界观测与输出反馈控制

许多实际系统可用n 维超球坐标系来描述, 并且系统有球对称的性质, 因而可通过研究半径方向的状态变化, 得到系统的全局动态过程. 通过将高维的对称系统转化为等价的径向一维方程, 本文采用边界Backstepping 方法设计了球对称反应扩散方程的输出反馈控制器. 使用容易测量的边界状态值, 设计了状态观测器来估计系统在空间域的所有状态, 从而实现输出反馈控制. 本文扩展了连续Backstepping 方法,提出了n维球坐标的Volterra 积分映射, 从而求出了显式表达的控制器和状态观测器. 论文用Lyapunov 函数法证明了输出反馈系统在H1范数下指数稳定, 表明状态对初值的连续依赖, 确保控制系统具有较好的性质, 不会在空间某点发散. 最后进行了数值仿真, 仿真结果表明系统在输出反馈控制律的作用下收敛到稳态值.