Recently, a lot of researches for lightweight materials and parts have been conducted to improve fuel efficiency in industry of transportation equipment such as aerospace and automotive. Especially, in order to reduce carbon-oxide gases and to improve fuel efficiency light-weighted metal parts have been applied to transportation equipment.
Aluminum alloys (Al-alloys) having high specific strength are effective for structural parts. So, Al-alloys have been used for structural frame parts related to safety of aircraft and automobiles. Also Al-alloys are applied to portable device demanding strength such as laptop computers and mobile phones for weight reduction. Especially Al-alloys are applied to parts having a thin-wall structure. A thin-wall structure is easy to be deformed by cutting forces, vibrations, and heat.
Chatter vibration generated by inappropriate cutting conditions makes surface integrity worse in thin-wall milling process of Al-alloys. Thus it is important to conduct continuous monitoring for diagnosis and avoidance of chatter vibration to ensure a machining quality in thin-wall milling process of Al-alloys.
The objective of this study is to analyze machining characteristics and to make a machining condition monitoring system to detect chatter vibration in thin-wall milling process of Al-alloy(Al7075-T651). For this purpose, thin-wall milling experiments are conducted to analyze the characteristics of machining process. Machining deformation and vibration characteristics of thin-wall milling process are identified by machined surface analysis and time-domain and frequency-domain analysis of cutting signals acquired by accelerometer, microphone, and AE sensor.
In addition, milling experiments are conducted to figure out detailed characteristics of chatter vibration which cause deterioration of surface integrity in thin-wall milling process of Al-alloy. Changes in cutting signals and surface conditions according to spindle speed and axial depth of cut are analyzed. Signal processing is conducted for extracting cutting signal characteristics when chatter vibration occurs. The limits of depth of cut that chatter vibration generates figures out by analysis of acceleration RMS, SPL, peak-count by DWT, and surface conditions. Also features of FFT graphs and frequency bands having high peak and high density when chatter vibration occurs are found out by FFT analysis.
Finally the monitoring algorithm applying artificial neural network(ANN) learned by the cutting signal patterns and fuzzy inference method using acceleration RMS and SPL is created to conduct the monitoring for diagnosis of chatter vibration in real-time. Patterning of cutting signals is conducted by using wavelet packet transform and vector normalizing to use for input pattern of pattern recognition algorithm by ANN. This suggested monitoring scheme can be applied to machining process effectively for detection of abnormal machining conditions such as chatter vibration.