The normal temperature corrosion of VC coating on the substrate of Cr12MoV prepared by TD process was tested in 5% NaCl aqueous solution, its surface morphologies and corrosion components after salt spray were observe...The normal temperature corrosion of VC coating on the substrate of Cr12MoV prepared by TD process was tested in 5% NaCl aqueous solution, its surface morphologies and corrosion components after salt spray were observed with SEM and EDS, respectively, and the effects of salt spray on micro-structures of VC coating were analyzed. Moreover, the invalidation mechanism of VC coating after salt spray and its effect on substrate material were discussed. The experimental results shown that the uniformity and integrity of VC coating surface are destroyed by salt spray for 120 h, a large number of the pits are produced on the coating surface, and the coating falls off, which speeds corrosion breakage of its substrate; the oxidated film on its surface becomes rougher, broken and discontinuous, and falls off easily, which reduce the ability of resistance salt spray; the failure modes of VC coating after salt spray are expressed with falling off of oxidated film, stress concentration and pore effect and so on, the corrosion breakage of oxidated film is the corrosion result of deoxidization corrosion from oxygen and HCl produced by NaCl and vapor.展开更多
以无水硼砂、FeV50和FeSi45为主要原料,通过TD和扩散处理在Cr12MoV基体表面制备了VC涂层。利用扫描电镜(SEM)、X射线衍射(XRD)、X射线光电子能谱(XPS)和能谱分析(EDS)等手段对VC的组织结构和化学元素分布进行了表征。用能谱面扫描分析...以无水硼砂、FeV50和FeSi45为主要原料,通过TD和扩散处理在Cr12MoV基体表面制备了VC涂层。利用扫描电镜(SEM)、X射线衍射(XRD)、X射线光电子能谱(XPS)和能谱分析(EDS)等手段对VC的组织结构和化学元素分布进行了表征。用能谱面扫描分析了VC涂层表面和界面化学元素分布,对其界面冶金结合机理进行了探讨。结果表明,VC涂层为C和V两元素组成的化合物,其中V原子浓度大约是C原子的2倍;VC涂层为单一的VC相,V元素电子结合能位于512.8 e V,C元素电子结合能位于282.2 eV;V、C和Fe等元素在涂层界面处发生了相互扩散,形成了界面扩散层,涂层-基体结合界面为冶金结合方式。展开更多
为适应实际大规模M arkov系统的需要,讨论M arkov决策过程(MDP)基于仿真的学习优化问题.根据定义式,建立性能势在平均和折扣性能准则下统一的即时差分公式,并利用一个神经元网络来表示性能势的估计值,导出参数TD(0)学习公式和算法,进行...为适应实际大规模M arkov系统的需要,讨论M arkov决策过程(MDP)基于仿真的学习优化问题.根据定义式,建立性能势在平均和折扣性能准则下统一的即时差分公式,并利用一个神经元网络来表示性能势的估计值,导出参数TD(0)学习公式和算法,进行逼近策略评估;然后,根据性能势的逼近值,通过逼近策略迭代来实现两种准则下统一的神经元动态规划(neuro-dynam ic programm ing,NDP)优化方法.研究结果适用于半M arkov决策过程,并通过一个数值例子,说明了文中的神经元策略迭代算法对两种准则都适用,验证了平均问题是折扣问题当折扣因子趋近于零时的极限情况.展开更多
基金Funded by the Foundation for Applied Basic Research of Changzhou City, China (CJ20110019)the Innovation Program of Graduated Student of Jiangsu Province(XM10-243)
文摘The normal temperature corrosion of VC coating on the substrate of Cr12MoV prepared by TD process was tested in 5% NaCl aqueous solution, its surface morphologies and corrosion components after salt spray were observed with SEM and EDS, respectively, and the effects of salt spray on micro-structures of VC coating were analyzed. Moreover, the invalidation mechanism of VC coating after salt spray and its effect on substrate material were discussed. The experimental results shown that the uniformity and integrity of VC coating surface are destroyed by salt spray for 120 h, a large number of the pits are produced on the coating surface, and the coating falls off, which speeds corrosion breakage of its substrate; the oxidated film on its surface becomes rougher, broken and discontinuous, and falls off easily, which reduce the ability of resistance salt spray; the failure modes of VC coating after salt spray are expressed with falling off of oxidated film, stress concentration and pore effect and so on, the corrosion breakage of oxidated film is the corrosion result of deoxidization corrosion from oxygen and HCl produced by NaCl and vapor.
基金The Science and Technology Support Project of Jiangsu Province,China(BE2014818)
文摘以无水硼砂、FeV50和FeSi45为主要原料,通过TD和扩散处理在Cr12MoV基体表面制备了VC涂层。利用扫描电镜(SEM)、X射线衍射(XRD)、X射线光电子能谱(XPS)和能谱分析(EDS)等手段对VC的组织结构和化学元素分布进行了表征。用能谱面扫描分析了VC涂层表面和界面化学元素分布,对其界面冶金结合机理进行了探讨。结果表明,VC涂层为C和V两元素组成的化合物,其中V原子浓度大约是C原子的2倍;VC涂层为单一的VC相,V元素电子结合能位于512.8 e V,C元素电子结合能位于282.2 eV;V、C和Fe等元素在涂层界面处发生了相互扩散,形成了界面扩散层,涂层-基体结合界面为冶金结合方式。
文摘为适应实际大规模M arkov系统的需要,讨论M arkov决策过程(MDP)基于仿真的学习优化问题.根据定义式,建立性能势在平均和折扣性能准则下统一的即时差分公式,并利用一个神经元网络来表示性能势的估计值,导出参数TD(0)学习公式和算法,进行逼近策略评估;然后,根据性能势的逼近值,通过逼近策略迭代来实现两种准则下统一的神经元动态规划(neuro-dynam ic programm ing,NDP)优化方法.研究结果适用于半M arkov决策过程,并通过一个数值例子,说明了文中的神经元策略迭代算法对两种准则都适用,验证了平均问题是折扣问题当折扣因子趋近于零时的极限情况.
