Monitoring of Spatters by Using Microphone during Gas Tungsten Arc Welding of Galvanized High-Strength Steel DP 980

摘要

Galvanized high-strength steels have broad applications in automobile industry due to their good formability, high strength and corrosion-resistant capability. Even with a reduced weight, a better crash performance can still be achieved. However, during the welding process of galvanized high strength steel, there are so many spatters produced due to a lower boiling temperature point of the zinc coat (907°C) than the melting temperature point of steel (1500°C). These spatters greatly deteriorate the quality of the welds, resulting in a disqualified product. Therefore, it is necessary to on-line monitor the presence of spatters; thus, achieving quality control. In this paper, a microphone was used to acquire the air-borne acoustic signals during gas tungsten arc welding (GTAW) of galvanized high-strength steel DP980 in an overlap-joint configuration. The acquired air-borne sound was denoised by spectral subtraction noise reduction method and analyzed in the wavelet domain. The results show that by applying noise reduction method, spatters can be well detected in a noisy environment due to their stronger sound pressure. In addition, the application of wavelet analysis has shown a good correlation between the denoised signals in the frequency range of [0:100 Hz] and the quality of the welds, which can be explained by the frequency characteristic of the oscillation behavior of the molten pool during arc welding. This paper studied the feasibility of using airborne sound signature to achieve on-line monitoring of the presence of spatters during the welding process of galvanized high-strength steels in a noisy environment by applying proper digital signal processing methods.