摘要 |
在自動化生產加工系統中,刀具狀態偵測系統扮演著極為重要的角色,當刀具發生磨耗或崩損時,適時的更換刀具能減少不良廢品造成的損失及提昇產品品質與生產良率。本研究之目的在於針對刀具狀態偵測系統開發進行初步探討及研究,根據許多文獻指出加工時震動訊號之特徵頻率能量與刀具狀態有密切之相關性。但為了分離出特徵頻率使用傅立葉轉換法須選擇適當頻段的濾波器,這對沒有相關專業知識背景的操作員而言是相當困難的,為解決此問題本研究提出一簡易量化頻譜變化的演算法-頻譜相關法,實驗結果顯示頻譜相關法能有效辨識出刀具磨耗及微崩的發生但無法得知刀具崩損之位置。對此本研究也針對切削力進行測量及探討希望藉由比對切削力之變化對應出刀具發生崩損之訊號門檻與可能位置,根據實驗結果顯示當刀具崩損位置切入工件後其切削力明顯劇增,故切削力本身即可作為辨識刀具損壞之指標,因此利用切削力之特徵趨勢配合閾值建立之演算法系統便能輕易達到辨識刀具微崩發生位置之目的,根據實驗結果顯示此法能有效的達到刀具微崩發生位置之辨識。
關鍵字:刀具狀態偵測、頻譜相關法、微崩偵測
It is well known that tool
condition monitoring system plays an important role in automatic
machining system. By detecting, and thus changing the worn tool
in time, the loss due to defect products can be greatly reduced
and hence ensuring product quality and reliability. The purpose
of this research is to develop a tool condition detection and
monitoring system for the tool wear and breakage during cutting
process. According to many research findings, the characteristic
frequency energy of cutting vibration signal gives the best
information corresponding to the tool condition. However, in
order to isolate the characteristic frequencies, it requires
selecting appropriate filters that are difficult for the
less-skilled operators in applications of the Fourier based
methods. To avoid this difficulty, spectrum correlation method
is proposed in this study. The experimental results showed that
the spectrum correlation method was able to detect the tool wear
and chipping, but it is unable to find out the chipping position
of the worn tool. For this reason, further investigation was
made to the cutting forces. The experimental results showed that
cutting forces increase sharply right after the tool chipping
zone was engaged into the workpiece. A rapid change in cutting
forces can itself be a good indicator to detect the tool
failure. According to the force features received, the tool
chipping detection and monitoring system, that has the capacity
to recognize the chipping position of cutting tool, was
developed. Experimental verification was conducted with a high
degree of success.
Key words: tool condition
monitoring, tool chipping detecting, spectrum correlation.
|