| Summary: | 碩士 === 國立臺灣大學 === 機械工程學研究所 === 102 === Machining efficiency and quality are usually affected by the status of the machine tool and cutting conditions. A suitable cutting condition for the work piece material is very important. In order to avoid these factors that cause significant damage, the processing monitoring must be done.
It’s relatively easy using a single type of sensors and signal processing during the development. However, we are monitoring the processing conditions which depend on the composition of the signal and certain physical phenomena and characteristics. To fully meet the actual situation, these signals must qualify to repeatability, reliability, responsiveness and resolution. In theory, the collection and analysis of multiple signals can get more comprehensive information. In short, sensor fusion is the solution of the demands above.
The purpose of this study is to use both acoustic emission signals and cutting temperature to monitor the tool condition. By experimenting turning of 1045 steel under different cutting conditions, we are able to establish the relation between signals and tool wear thus can be extended in the case of different cutting conditions, and can estimate flank wear accurately. According to the results of this study, we can estimate if the tool wear level were within the range of appropriate processing or not with some simple cutting experiments, which provide information clearly and are easy to identify.
The use of sensor fusion concepts, using acoustic emission signals and temperature detection, can identify how the acoustic emission signals, temperature, cutting speed, and wear were related, then establish mathematical models to calculate the wear level. We can estimate wear level through fewer experiments, particularly the integration of sensor fusion can effectively eliminate the impact of circumstances or inaccurate sensor failure in particular reasons caused situations.
|
|---|
