Nowadays high-strength steels have great applications in different industries due to their good combination of formability, weldability, and high strength-to-weight ratio. To guarantee a high quality without the presence of defects such as partial penetration (PP) in the laser welding of high-strength steels, it is very important to on-line monitor the whole welding process. While optical sensors are widely applied to monitor the laser welding process, we are proposing to use a microphone to acquire the airborne acoustic signals produced during laser welding of high-strength steel DP980. In order to extract valuable information from a very noisy signal acquired in a harsh environment such as industrial welding, spectral subtraction (SS), a noise reduction method is used to process the acquired airborne sound signals. Furthermore, by applying the power spectrum density (PSD) estimation method, the frequency characteristics of the acoustic signals are analyzed as well. The results indicate that the welds in full penetration (FP) and PP produce different signatures of acoustic signals that are characterized with different sound pressure levels and frequency distributions ranging from 500 Hz to 1500 Hz. Based on these differences, two algorithms are developed to distinguish the FP from PP during the laser welding process. A real-time monitoring system is implemented by a LabVIEW-based graphic program developed in this research. A feedback control system that could guarantee the FP will be developed in the near future.
