Heat Transfer Interface Guided Design for an Instrument to Measure the Heat Transfer Coefficient in Friction Stir Welding

Friction stir welding (FSW) is a solid-state joining process that offers superior weld quality compared to traditional fusion welding, with specific advantages in aerospace, automotive, and nuclear fields and materials. Important process parameters include the translational velocity, rotation rate, and downward force, all of which affect the temperature fields within the welded material and thus the weld quality. Important parameters that cannot be directly controlled include the friction coefficient and heat transfer coefficient between the tool head and welded material. Current literature presents the heat transfer coefficient within a range spanning up to three orders of magnitude. The temperature fields, and thus weld quality, are strongly influenced by the heat transfer coefficient. Because of the complex nature of FSW processes and the uncertainty of the friction and heat transfer coefficients, FSW processes have typically been limited to trial-and-error approaches to fine-tune the welding parameters within satisfactory bounds. The goal of the present research is to develop and validate an instrument design to directly and accurately measure the heat transfer coefficient between the tool head and welded material. This would provide an accurate heat transfer coefficient, reducing the parameters needing to be tuned to just the friction coefficient. With just one parameter to tune, the finite element analysis (FEA) models will be able to quickly converge on accurate parameters. The device is based on the frequency domain thermoreflectance (FDTR) technique to locally heat and sense the temperature of an insert that is designed to fit within the FSW tool. An analytical model was then developed for the FDTR system geometry using a thermal quadrupoles model. The model determines the phase delay between pump and probe lasers as a function of frequency and heat transfer coefficient, resulting in phase delay curves. A curve fit can then be applied to experimental data to determine the heat transfer coefficient. COMSOL was initially used to validate these analytical expressions for simple multi-layered geometries. After these expressions were validated, they were used to perform a sensitivity analysis on potential geometries for use in a FSW environment. Sensitive parameters were maximized, resulting in a final optimal configuration. COMSOL was then used for another series of tests, confirming that the optimal configuration matched the analytical expressions within the radial constraints of the FSW configuration. The analytical expressions and COMSOL results match well in the instance of simple static geometry. The results give evidence to support their functionality and suitability in a real FSW environment to determine the heat transfer coefficient. This work will result in a novel instrument that can directly measure the heat transfer coefficient in situations previously not done, specifically enabling rapid FSW parameter tuning. This will result in significant monetary savings from the reduced number of process tests to fine tune parameters. This instrument will also be able to be used to measure variations in the local heat transfer coefficient with greater resolution than ever before, providing insight into FSW mechanics.