课程文献中心 更多>>

关键词： 铀铌合金   神经网络势函数   分子动力学   力学性能  

摘要： 铀铌合金在不同实验环境中呈现出复杂的晶体相和独特的力学性能,但原子尺度的相析出和形变损伤机制尚不清楚,其根本原因是缺乏支撑大尺度分子动力学模拟的精确铀铌合金原子相互作用势.本工作基于自主开发的神经网络势能函数及随机搜索方法,构建了覆盖全化学空间的铀铌合金第一性原理计算数据库,并基于神经网络框架建立了具有较高泛化性能和精度的铀铌二元体系机器学习势函数,其能量和力的测试平均绝对误差分别为5.6 meV/atom和0.095 eV/Å,可以精确地描述不同化学成分铀铌合金的晶体空间结构、状态方程及热力学参量.基于该势函数,我们实现了低温时效下铀铌合金相失稳分解过程的原子尺度模拟,初步阐明了 Nb析出相对其合金力学性能的影响及原子响应机制.

关键词： Education policy   mobility   network   policy borrowing   future education   global governance  

摘要： The global policy discourse surrounding future education has propelled reform movements worldwide. This study explores the policy mobilities and assemblage of New Pedagogies for Deep Learning (NPDL), an edu-business that has rapidly expanded to promote its distinct future learning models. Specifically, it investigates the adaptation of NPDL policy within Taiwan, while concurrently exploring the service provided by the NPDL global network, and the transnational policy learning between Taiwan and Australia. First, we analyse the integration of NPDL into Taiwan's national reform agendas by tracing the interactions amongst diverse actors and organisations. Second, we explore the policy recontextualisation process by following Taiwanese policy actors as they travel to Sydney to engage with their Australian counterparts. A central argument is that future education policy is shaped beyond national borders and continuously evolves through the interplay of (sub)national, transnational, and global actors, where ideas and practices are reconfigured as they move across spaces.