工具機主軸品質診斷系統開發

Development of Fault Diagnosis System For High Speed Spindle

謝男凱(Nan-Kai Hsieh)

指導教授 ： 楊宏智

國立臺灣大學/工學院/機械工程學研究所/博士(2015年)

https://doi.org/10.6342/NTU.2015.11053

摘要

隨著自動化加工技術的蓬勃發展，加工設備亦延長運作時間。當設備長時間運作時，所產生之振動與噪音，勢必對設備造成磨損與故障等狀況發生，進而影響生產線的稼動率。目前受限於檢測方法與設備能力，往往無法預知故障之發生。假若能有效偵測故障即將發生，便能將設備故障所帶來之損失降之最低。因此，近年來許多智能監控與無人化工廠廣泛運用於製造業之產線上。本研究中，乃是根據被喻為工具機心臟的高速主軸，進行智能化高速主軸系統的開發。首先，透過製程管系統(Manufacturing Execution System, MES)收集了高速主軸之生產與維修資訊，且整理出常見之故障原因。再利用定壓預壓主軸，模擬出每種損壞狀況下之振動情形，並收集三軸向之振動訊號。常見之高速主軸檢測方法，乃是將振動訊號經由傅立葉轉換，再透過頻譜圖觀察不同頻率下之振幅值；本文並加入多尺度熵做為損壞辨識方法，藉由計算訊號在各尺度下的亂度值，做為判斷方式。
本研究最後透過重複性實驗擷取損壞模型之振動訊號，並搭配演算法找出各損壞模型之閥值。將各損壞模型之閥值彙整後，匯入所開發之損壞辨識系統(Defect Diagnosis System)，此套系統能針對不同之損壞狀況，以即時訊息之方式告知操作人員，並將系統實際應用於高速主軸進行跑合前之故障檢測。其主要目的為，高速主軸可透過跑合前的篩檢，預先檢測出原先振動計所無法辨識出之潛在故障因子，有效防止主軸在跑合過程中發生軸承損毀之狀況。目前損壞辨識系統已能實際應用於生產線上，做為產品品管之工具，且辨識準確率已能達9成以上。

關鍵字

高速主軸 ； 振動訊號 ； 傅立葉轉換 ； 多尺度熵 ； 損壞辨識

並列摘要

During the long machining cycles, the vibration and accompanied noise produced by the machine tool may cause unduly mechanical wear and, consequently malfunction. The inadequate facility limits an effective malfunction pre-warning diagnosis and this often leads to the sudden failure of the spindle. Therefore, it is necessary to rely on the real-time monitor system to enhance the reliability of the machine tool. This research collects the information in the process of production and maintenance, from which the common faults are summarized, and then a constant-preload spindle is used to collect vibration signals and simulate each failure mode. This research also proposes the multi-scale entropy (MSE), in addition to the most common detection of vibration signals, the Fourier analysis. It is used to observe the amplitude of each frequency at each instant. The MSE can be used to calculate the MSE curve which can then be used to correctly identify some defect modes. In order to prevent the failure without warning, this research is designed to extract the threshold of each fault condition by repetitive experiments to develop the defect diagnosis system (DDS).
The main objective of the DDS is applied to detect the faults before the run-in period of spindle so that any potential errors can be identified and recognized. These potential damaging factors are detected from the spindle to avoid breaking the bearing during the run-in period. At present, the DDS is applied successfully in the production line for quality control. and the recognition rate of DDS is high up to 90 percent
.

並列關鍵字

High Speed Spindle ； Vibration Signal ； Fourier Transform ； Multi-scale entropy ； Defect Diagnosis System

參考文獻

[48] 林志明, “滾珠軸承之剛度分析,” 碩士論文, 中原大學機械工程研究所, 2002

[1] Guo-Hua Feng, Yi-Lu Pan, “Investigation of Ball Screw Preload Variation Based on Dynamic Modeling of a Preload Adjustable Feed-Drive System and Spectrum Analysis of Ball-Nuts Sensed Vibration Signals,” International Journal of Machine Tools & Manufacture, 52, (2012) pp. 85-96

[2] Yaguo Lei, Zhengjia He, Yanyang Zi, “Application of an Intelligent Classification Method to Mechanical Fault Diagnosis,” Expert Systems with Applications, 36, (2009) pp. 9941-9948

[3] Dongyang Dou, Jianguo Yang, Jiongtian Liu, Yingkai Zhao, “A Ruled-Based Intelligent Method for Fault Diagnosis of Rotating Machinery,” Knowledge-Based Systems, 36, (2012) pp. 1-8

[4] M. Amarnath, I.R. Praveen Krishna, “Detection and Diagnosis of Surface Wear Failure in a Spur Geared System using EEMD Based Vibration Signal Analysis,” Tribology International, 61, (2013) pp. 224-234