| Summary: | This thesis is to investigate the machining surface roughness by using acoustic emission (AE) method. The objectives of this project is to acquire AE data of the experiment by operating milling process, to study the correlation of AE parameter with work piece surface roughness (R) and to cluster AE data by using time domain analysis such as global statistical parameter and clustering method for online machining condition monitoring. In order to done this experiment, there is method to be taken. Firstly is the experimental setup. Computational Numerical Control (CNC) milling machine will be use through this project conduct the face milling. Machining parameter set for depth of cut, cutting speed and feed rate. Surface roughness being measure by using perthometer. USBwin for AE Node used for data acquisition. The material used is Hayness 188 alloy and carbide-coated as the cutting tool. Before experiment is started, the AE system need to be tested by using pencil break test to check whether AE system can receive AE signals properly. When lead of pencil break, it will generate as equal as AE signals emit during experiment. For data analysis, AE signal can be cluster based on surface roughness. For clustering analysis, it is related to its signal properties. Method used to cluster the signals is global statistical parameter such as root mean square (RMS), skewnes, kurtosis, peak value and variance. Based on experiment data, the pattern of AE parameter with time domain analysis can be concluded by clustering method. The analysis shows that AE signals data can be cluster by global statistical parameter according to its surface roughness measurement. Between all global statistical parameter, we can see that peak value, RMS and variance can show the significant pattern on clustering. As the project is success, data collected surface roughness monitoring can be made and be used in industry. So, this method can be use as an alternative method in industry to decrease the time used and the cost needed.
|
|---|