Optical plasma spectroscopy as a tool for monitoring laser welding processes

In this work we present experimental results of the joint application of the two previously mentioned spectroscopic techniques (electron temperature and correlation analysis) to a real-time laser welding monitoring case study. The two signals have been calculated starting from selected chemical species composing the plasma spectra. Experimental evidence is given of the correlation between the recorded signals and the occurrence of weld defects intentionally generated by varying the laser power and the travel speed. An optical sensor prototype was used, that embeds a fiber-coupled miniature spectrometer having a dynamic spectral range from 390 nm to 580 nm and a resolution of 0.3 nm. Such a prototype employed data acquisition and real-time spectra analysis algorithms for both the previously mentioned spectroscopic techniques, e.g. the electron temperature and the correlation coefficients. A high power CO2 laser source was used with maximum output power of 6 kW. The laser-metal interaction zone was shielded by an argon flow. The plasma optical emission was collected by a quartz collimator and transmitted to the optical sensor by a 50 μm core-diameter optical fiber. Spectral lines from three different chemical species (Mn(I), Fe(I), Cr(I)) composing the plasma plume and the stainless steel alloy were used for the acquisition of the electron temperature and the correlation signals. Compared to other optical sensors, the main advantage of this system is that it has a great flexibility upon variation of the welding metal or the joint geometries. In fact once the chemical composition of the alloy was known and most plasma emission lines are identified, only a slight calibration of the software settings is necessary.