Real-Time Detection of Burn-Through and Porosity in GTAW Using Ultrasonic and Acoustic Emission

2019-02-01T00:00:00Z (GMT) by Alexandra-Del-Carmen Basantes-Defaz
It is quite challenging to assess weld quality real-time and doing it accurately. A proper real-time nondestructive evaluation approach should be considered as an essential component of an automated weld process to produce quality welds consistently. Part of this study focuses on determining if NDE techniques, specifically acoustic emission (AE) and ultrasonic testing (UT) can be used to detect burn-through, weld porosity, and assesses weld penetration during weld fabrication. This thesis presents the development of a real-time welding defect detection system for gas tungsten arc welding (GTAW) and the challenge to integrate non-contact acoustic and ultrasonic methods to assess weld quality concerning the welding parameters (i.e., weld current, voltage, travel speed, and gas flow rate). The in-situ NDE monitoring system was used to evaluate the changes in weld morphology, burn-through by varying the weld heat input and assessing the presence of porosity by adding oxygen to the Ar-shielding gas. Different categories of burn-through were defined that includes the onset of burn-through with the back side of the base plate undergoing melting also with weld contained within the plate, and the production of real burn-through where a through thickness a cavity is produced in the base metal. The objective of this investigation was to evaluate real-time burn-through and porosity in gas tungsten arc welding (GTAW) with a combination of acoustic emission and ultrasonic NDE methods. Weld heat input and a mixture of Ar gas and Ar- 2%O2were the controlling parameters used to introduce burn through and porosity in the weld respectively. The AE data provides qualitative information about the weld condition and characteristics. The UT data provide more quantitative information about the weld morphology. Results demonstrated that complete air-coupled UT could not be used simultaneously with welding due to the influence of the magnetic field that develops in the welding torch during when the arc was struck, this resulted in the UT signal being eliminated. Consequently, a rolling UT transmitter was combined with an air-coupled UT receiver to increase the signal/noise value and able to assess the weld online. Wave dispersion due to the different levels of burn-through was detected. The AE signal was correlated with the heat input, and it was determined that surface discontinuities resulted in sudden surges in AE energy indicating that surface weld defects could be detected during processing. In conclusion, passive and active NDE methods could be combined to monitor weld quality real-time for assessment of the weld.