Signature Analysis And Tone Channel Generation Mechanism Of Pipeline MIG Welding Arc Sound

Metal inert gas welding (MIG) is widely used in pipeline welding. In the welding process, there is a lot of information about welding state in the arc sound. if extract and make use of this information in the pipeline welding, it will effectively improve the quality of welding. In this paper, having arc sound signal for the object, the relationship between MIG welding arc sound of pipe and weld penetration state and the sound of channel formation mechanism are related researched. The main contents as follows:Establishing a pipeline MIG welding experiments of hardware system platform and the software systems incorporating a series of arc acoustic signal collection, analysis and noise reduction. This system can collect double channel signal in real-time aimed at pipe welding arc sound, simultaneously collecting different locations of the arc acoustic signal at the welding process.On the acquisition pipeline welding arc sound signal makes noise reduction. Signal denoising selects 4 level wavelet function. Experimental results show that the noise in the signal is effectively filtered, and retain the mutation characteristics of the original signal. Signals become clearer, greatly increasing the readability. Effect of denoising is good. From the time domain, frequency domain and time-frequency domain three aspects researching the intrinsic link between the welding arc sound signals and the penetration state.The arc sound energy of external piping is mainly in the frequency range 4 to 10kHz to reflect the changing of penetration state, but at the internal piping mainly reflecting the penetration state change of the frequency range is from 0 to 4kHz. The External piping frequency band of reflecting the penetration is higher than the internal piping.It is discussed about the production of pipe welding arc sound mechanism and channel formation mechanism. The excitation source of piping arc sound is that plasma oscillation makes the surrounding media occur mechanical vibrations. This mechanical vibration spread as sound wave. Pipe welding pipe channel system is formed by its own relatively closed space and an oval cavity arc that exist in the arc. The interaction between excitation and the channel forms the arc sound. the resonant frequency of the channel determines pipeline arc sound characteristics spectrum. Pipeline arc sound model is established by linear prediction analysis (LPC). With the help of LabVIEW arc sound signals can be simulated. Experiments show that, LPC frequency response can be a good description of pipeline welding arc sound signal frequency spectrum envelope.