Actual heat treatment processes must face increasing specifications with reference to process quality, safety and results in terms of reproducibility and repeatability. They can be met only if the parts’ surface cond...Actual heat treatment processes must face increasing specifications with reference to process quality, safety and results in terms of reproducibility and repeatability. They can be met only if the parts’ surface condition is controlled during manufacturing and, especially, prior to the treatment. An electrochemical method for the detection of a steel part’s surface condition is presented, together with results, consequences, and mechanisms concerning surface pre-treatment before the thermochemical process. A steel surface’s activity or passivity can be detected electrochenucalry, independently from the chemical background. The selected method was the recording of potential vs. time curves at small constant currents, using a miniaturized electrochemical cell, a (nearly) non-destructive electrolyte and a potentio-galvanostatic setup. The method enables to distinguish types of surface contamination which do not interfere with the thermochemical process, from passive layers which do and must be removed. Whereas some types of passive layers can be removed using conventional cleaning processes and agents, others are so stable that their effects can only be overcome by applying an additional activation pre-treatment, e.g. oxynitriding.展开更多
针对水电机组状态监测数据量逐步增大,数据质量差的问题,提出了一种基于改进K维树(K-Dimensional Tree,KD-Tree)与基于密度的空间聚类算法(Density-Based Spatial Clustering of Applications with Noise,DBSCAN)的水电机组状态监测数...针对水电机组状态监测数据量逐步增大,数据质量差的问题,提出了一种基于改进K维树(K-Dimensional Tree,KD-Tree)与基于密度的空间聚类算法(Density-Based Spatial Clustering of Applications with Noise,DBSCAN)的水电机组状态监测数据清洗方法,首先对输入数据建立KD-Tree,再使用DBSCAN在最近邻样本上扫描完成聚类,聚类结束以后会分离出噪声点,将噪声点去除即可完成对水电机组状态监测数据清洗。选取某水电站状态监测系统上导摆度数据1 088条,再以相同时间间隔插入随机数据100条,通过算例与常规DBScan、K-means、OCSVM算法对比聚类性能与时间性能,所提出的方法识别正确率最高,为97.78%,消耗时间最少,为0.007 732 s,数据清洗效果最优,并可以大幅减少计算时间。展开更多
基金the financial support from AiF(Arbeitsgemeinschaft industieller Forschungsvereinigungen Otto von Guericke)and DFG(Deutsche Forschungsgemeinschaft)which made the work documented in this text possible.
文摘Actual heat treatment processes must face increasing specifications with reference to process quality, safety and results in terms of reproducibility and repeatability. They can be met only if the parts’ surface condition is controlled during manufacturing and, especially, prior to the treatment. An electrochemical method for the detection of a steel part’s surface condition is presented, together with results, consequences, and mechanisms concerning surface pre-treatment before the thermochemical process. A steel surface’s activity or passivity can be detected electrochenucalry, independently from the chemical background. The selected method was the recording of potential vs. time curves at small constant currents, using a miniaturized electrochemical cell, a (nearly) non-destructive electrolyte and a potentio-galvanostatic setup. The method enables to distinguish types of surface contamination which do not interfere with the thermochemical process, from passive layers which do and must be removed. Whereas some types of passive layers can be removed using conventional cleaning processes and agents, others are so stable that their effects can only be overcome by applying an additional activation pre-treatment, e.g. oxynitriding.
