The article raises the question of what to do with one of the main achievements of metal science in recent years—binary phase diagrams. These diagrams play a key role in the science of alloys and therefore their reli...The article raises the question of what to do with one of the main achievements of metal science in recent years—binary phase diagrams. These diagrams play a key role in the science of alloys and therefore their reliability must be complete. However, the discovery of the “ordering-separation” phase transition, which showed that in binary alloys at certain temperatures the sign of the chemical interatomic interaction changes (and, consequently, the microstructure changes), forces us to reconsider our ideas about those areas. Currently, these areas are designated on diagrams as areas of a “disordered solid solution.” This article proposes, using transmission electron microscopy, to study all the so-called solid solution regions, and apply the results obtained to the studied regions of the phase diagram.展开更多
文摘The article raises the question of what to do with one of the main achievements of metal science in recent years—binary phase diagrams. These diagrams play a key role in the science of alloys and therefore their reliability must be complete. However, the discovery of the “ordering-separation” phase transition, which showed that in binary alloys at certain temperatures the sign of the chemical interatomic interaction changes (and, consequently, the microstructure changes), forces us to reconsider our ideas about those areas. Currently, these areas are designated on diagrams as areas of a “disordered solid solution.” This article proposes, using transmission electron microscopy, to study all the so-called solid solution regions, and apply the results obtained to the studied regions of the phase diagram.
文摘针对现有基于数据驱动的随机子空间(data-driven stochastic subspace identification,DATA-SSI)算法存在的不足,无法实现稳定图中真假模态的智能化筛选,提出了一种新的模态参数智能化识别算法。首先通过引入滑窗技术来实现对输入信号的合理划分,以避免虚假模态和模态遗漏现象的出现;其次通过引入OPTICS(ordering points to identify the clustering structure)密度聚类算法实现稳定图中真实模态的智能化筛选,最后将所提算法运用于某实际大型斜拉桥主梁结构的频率和模态振型识别过程中。结果表明,所提改进算法识别的频率值结果与理论值(MIDAS有限元结果)以及实际值(现场动力特性实测结果)间的误差均在5%以内,且识别的模态振型图与理论模态振型图具有很高的相似性。
