Tool wear evaluation under minimum quantity lubrication by clustering energy of acoustic emission burst signals

Abstract

Tool wear evaluation by Acoustic Emission (AE) is a promising non-destructive technique without stopping the machine, which can save its off-line detection time and prevent tool re-clamping error. However, the traditional tool wear evaluated by AE continuous signals Root Mean Square (RMS) is inaccurate and relative difficult especially when machining ductile material, not to mention it’s evaluation under the Minimum Quantity Lubrication (MQL). Thus, this paper proposes a non-destructive method of tool wear evaluation by clustering energy of AE burst signals under MQL cutting condition. The interface AE burst signals are firstly eliminated by properly setting the AE parameters including threshold, rise time and duration time. Then, AE burst signals are identified to three categories, which are induced by MQL, material fracture and plastic deformation. Finally, the relationship between tool flank wear and AE burst signals are established by clustering AE energy. The results show that tool flank wear can be accurately evaluated by linear fitting total energy of AE burst signals induced by fracture and plastic deformation.