Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Grinding burn is a common constraint to grinding operations. Grinding burn damages materials and degrades properties of materials, because it causes tensile residual stresses or micro-fractures in the workpiece surface. The main cause of grinding burn is high grinding temperature. The main problems of current method for grinding burn identification are their sensitivity and robustness. This paper presents a successful identification of grinding burn with very high sensitivity acoustic emission (AE) monitoring. Experimental results show that the accuracy of grinding burn recognition can reach 92% at its best. Grinding AE signals contains rich information for grinding process monitoring. A successful grinding process monitoring relies on the effective feature extraction from complex AE signals. A method of grinding burn detection using AE information has been represented in this paper. A number of AE features were extracted by wavelet packet transform and the most characteristic features were selected by fuzzy clustering identification. It has demonstrated that the wavelet packet transform can capture features that are sensitive to grinding burn, but insensitive to the variations of grinding conditions. Fuzzy clustering analysis can get rid of redundant features, which may cause inaccuracy of the grinding monitoring system. By means of fuzzy pattern recognition, grinding burn can be identified based on distance criteria. Experiment results demonstrate a successful application in grinding burn monitoring.
Grinding burn is a common constraint to grinding operations. Grinding burn damages materials and degrades properties of materials, because it causes tensile residual stresses or micro-fractures in the workpiece surface. The main cause of grinding burn is high grinding temperature. The main problems of current method for grinding burn identification are their sensitivity and robustness. This paper presents a successful identification of grinding burn with very high sensitivity acoustic emission (AE) monitoring. Experimental results show that the accuracy of grinding burn recognition can reach 92% at its best. Grinding AE signals contains rich information for grinding process monitoring. A successful grinding process monitoring relies on the effective feature extraction from complex AE signals. A method of grinding burn detection using AE information has been represented in this paper. A number of AE features were extracted by wavelet packet transform and the most characteristic features were selected by fuzzy clustering identification. It has demonstrated that the wavelet packet transform can capture features that are sensitive to grinding burn, but insensitive to the variations of grinding conditions. Fuzzy clustering analysis can get rid of redundant features, which may cause inaccuracy of the grinding monitoring system. By means of fuzzy pattern recognition, grinding burn can be identified based on distance criteria. Experiment results demonstrate a successful application in grinding burn monitoring.
Grinding burn identification through the AE monitoring
Identifizierung eines Schleifbrennens durch Schallemissionsüberwachung
2004
9 Seiten, 5 Bilder, 4 Tabellen, 17 Quellen
Aufsatz (Konferenz)
Englisch
Grinding Burn Mechanism of Directionally Solidified Superalloy
| British Library Online Contents | 2001
Burn threshold prediction for high efficiency deep grinding
| Tema Archiv | 2011
Grinding Burn and Chatter Classification Using Genetic Programming
| British Library Online Contents | 2009
Signal Analysis of Acoustic Emission for Laser Imitating Grinding Burn
| British Library Online Contents | 2005
Temperature Measurement and Burn Mechanism of Stainless Steel 1Cr11Ni2W2MoV in Grinding
| British Library Online Contents | 2012