Microstructure and high-temperature corrosion behaviors of aluminide coatings by low-temperature pack aluminizing process

Aluminide coatings were produced on carbon steel and Fe-5Cr-Mo alloy at a relatively lower temperature below 600 °C in shorter treatment time by a combination of surface refinement process and pack aluminizing process. Repetitive ball impact, generated by mechanical vibration, caused the top-layer refinement of substrates in a conventional pack aluminizing process. The effects of temperature and treatment time on the formation of aluminide coatings were analyzed. The microstructure of the coatings was investigated by SEM, AFM and XRD. The aluminide coatings were one-layer, compacted structure with ultrafine grains and uniform elemental distribution. High-temperature oxidation and sulphidation tests were carried out at 600 °C in air for 200 h and 10% SO₂ + Ar gas mixture atmosphere for 50 h, respectively. The mass gains and spallation indicated that the aluminide coatings significantly improved the high-temperature oxidation and sulphidation resistance.     

Absolute small-angle measurement based on optical feedback interferometry

We present a simple but effective method for small-angle measurement based on optical feedback interferometry （or laser self-mixing interferometry）. The absolute zero angle can be defined at the biggest fringe amplitude point, so this method can also achieve absolute angle measurement. In order to verify the method, we construct an angle measurement system. The Fourier-transform method is used to analysis the interference signal. Rotation angles are experimentally measured with a resolution of 10^-6 rad and a measurement range of approximately from -0.0007 to ＋0.0007 rad.     

Highly enhanced luminescence of ZnO green phosphors by calcining ZnS with NH4Br as additive

Zinc oxide (ZnO) phosphors with highly efficient green emission have been prepared by calcining ZnS with NH₄Br as additive in air atmosphere. The luminescent properties of as-prepared ZnO phosphors were characterized by X-ray photoelectron spectroscopy and photoluminescence. Our results reveal that the green emission is ascribed to a transition of a photo-generated electron from the localized defect centers (V_o⁺) to a deeply trapped hole (V_Zn⁻) within the band gap. The addition of NH₄Br enhances the luminescent emission of ZnO by promoting the formation of vacancies of both oxygen and zinc.     

图形处理器并行计算用于离子发动机粒子模拟

 为了研究离子发动机羽流对航天器的影响,采用质点网格-蒙特卡罗碰撞方法对离子发动机羽流中的交换电荷离子进行了模拟。利用计算设备统一架构技术,开发出一套基于图形处理器的并行粒子模拟程序。随机数生成采用并行MT19937伪随机数生成器算法,电场方程使用完全近似存储格式的代数多重网格法求解。r-z轴对称坐标系中,在z=0 m处获得的电流密度均值为4.5×10^-5 A/m²,图形处理器所得结果与中央处理器模拟结果吻合。在16核心的NVIDIA GeForce 9400 GT图形显示卡上,取得相对于Intel Core 2 E6300中央处理器4.5~10.0倍的加速比。     

A new solvent suppression method via radiation damping effect

Radiation damping effects induced by the dominated solvent in a solution sample can be applied to suppress the solvent signal. The precession pathway and rate back to equilibrium state between solute and solvent spins are different under radiation damping. In this paper, a series of pulse sequences using radiation damping were designed for the solvent suppression in nuclear magnetic resonance (NMR) spectroscopy. Compared to the WATERGATE method, the solute signals adjacent to the solvent would not be influenced by using the radiation damping method. The one-dimensional (1D) ¹H NMR, two-dimensional (2D) gCOSY, and J-resolved experimental results show the practicability of solvent suppression via radiation damping effects in 1D and 2D NMR spectroscopy.     

自动波竞争神经网络及其在单源最短路问题中的应用

将竞争机理引入网络的自动波产生与传播过程中，提出自动波竞争神经网络(ACNN)模型，并成功地应用于求解单源最短路问题，给出了基于ACNN的最短路求解算法. 与其他神经网络最短路算法相比，基于ACNN的最短路算法具有网络所需的神经元数目少、神经元及网络的结构简单、易于软硬件的实现、以及全并行方式计算等特点，可用于求解非对称赋权图的最短路径树问题，且其计算时间(迭代次数)仅正比于最短路径上的跃点数，与赋权图的复杂度、路径总长、边长的精度、通路总数等因素无关. 计算机仿真结果表明该算法的有效性和快速求解能力. 关键词： 单源最短路问题 自动波竞争神经网络 脉冲耦合神经网络     

液面法超声全息成象的理论分析与实验结果

本文采用傅里叶分析方法,对液面法超声全息成象原理作了较为严格的理论分析,给出了物体束的空间分布为任意函数时,液面起伏的计算公式.并给出了物体束为空间方脉冲时,液面起伏、重现之一级象（即通常称为全息象）,零级象（相应于普通的非全息象）的计算结果,讨论了影响象的强度、畸变、分辨率的因素等问题,对全息象和非全息象的优缺点作了比较。最后,还给出了部分实验结果,其结果与理论计算比较符合。     

基于人工神经网络的商业银行信用风险模型

在对人工神经网络的基本原理进行简要介绍的基础上 ,着重对构建商业银行信用风险的人工神经网络模型进行了研究 ,实证结果表明 ,人工神经网模型具有很高的预测精度