文章詳目資料

技術學刊 EIScopus

  • 加入收藏
  • 下載文章
篇名 軸彎與不平衡位置未知之快速故障診斷方法
卷期 25:1
作者 楊大中徐銘韋
頁次 37-48
關鍵字 故障診斷旋轉機械軸彎不平衡fault diagnosisrotating machineryshaft bowimbalanceEIScopusTSCI
出刊日期 201003

中文摘要

大型旋轉機械 (如電廠燃氣渦輪機與蒸汽渦輪機等) 之轉子系統長期運轉,常常會因葉片遭受侵蝕或積垢,造成不平衡現象,及因製造公差或缺陷、重力效應造成永久性沉陷與表面受熱分布不均,可能導致軸彎。這些故障可能是分散在多處,其產生可能是漸進累積的,不易由運轉中之機械振動訊號診斷出來。本文提出一診斷方法可以在固定轉速下,在事先未知故障數量與位置之
狀況下,經由軸承處之振動訊號,快速鑑別軸彎及不平衡的發生位置與故障量。過去同時鑑別軸彎及不平衡的方法需要低速與高速之轉軸振動訊號,才可以分辨軸彎與不平衡,本文的特點即在固定轉速下的訊號即可同時鑑別軸彎與不平衡。過去鑑別故障的方法需事先知道故障數量與位置,本文方法並無此限制,可在事先未知故障數量與位置之狀況下進行故障診斷。本方法之所以能夠迅速鑑別的原因,在有秩序計算可能節點,省去隨機選擇節點的計算時間。由以上優勢,本文方法可以應用在固定轉速之大型旋轉機械 (例如電廠汽渦輪發電機組,固定轉速3600 rpm) 之故障診斷與線上監控,掌控系統運作之狀況。

英文摘要

Large rotating machines, like gas turbines and steam turbines in power
plants, may suffer from the faults of imbalance, caused by erosion and
deposition on blades, and shaft bow, caused by sagging due to manufactur-
ing tolerance or defects, and/or sagging due to gravity or uneven thermal
distribution. These faults, imbalance and shaft bow, are difficult to
distinguish because they may be distributed over multiple locations and
develop gradually during a long-term period of operation. A fast fault
diagnostic method is proposed to identify the numbers and locations of
faults based on the vibration responses of the rotor, without prior
knowledge of the numbers and locations of the faults.
Previous diagnosis methods have required vibration responses of
rotors at low and high rotating speeds to distinguish the faults of shaft bow and imbalance while only the vibration responses of the rotor at a fixed rotating speed is needed in the proposed method. Most diagnosis methods identify the faults of known numbers and locations, while the proposed method can identify faults under conditions of not knowing the numbers and locations of the faults in advance. Compared with other fault diagnosis methods that search all the possible locations of faults and consume a large amount of computation power and time, the proposed
method achieves fast diagnosis by orderly searching through the possible
locations of faults and using far less computer time and power than other
methods. Based on the above advantages, the proposed method is suitable for the online diagnosis and monitoring of large rotating machinery of fixed rotating speeds, like steam turbine-generators in power plants, to monitor the health of the machines.

相關文獻