A metal-cutting-inspired analytical framework for predicting operating parameters in friction stir welding
Date of Award
Doctor of Philosophy (PhD)
Mechanical and Aerospace Engineering
Friction stir welding, Metal-cutting
Currently, predictive models for establishing Friction Stir Welding (FSW) process parameters that produce repeatable and robust welds are not well developed and do not account for differences between alloys. Parameter development relies primarily upon costly trial and error studies. The predicted hot working conditions of FSW vary widely with strains ranging from 2 to 100 and strain rates from 10^0 to 10^6 s^-1. These conditions lie outside the range covered by traditional testing techniques but are quite similar to conditions predicted in Orthogonal Metal Cutting (OMC), where strains range from 1 to 50 and strain rates from 10^0 to 10^6 s^-1. Additionally, the shear interface in FSW is similar to that of OMC. In this work, a simplified analytical model of the FSW process was derived from the OMC process, calibrated with experimental data in AA1100, AA2219, and AA6061, and used to predict parameter windows in AA5083 and AA7075. This model is capable of predicting hot working conditions such as strain, strain rate, flow zone geometry, temperature and flow stress based on material properties and FSW process parameters such as tool rotational and translational velocities. Parameters were evaluated using a flow stress criteria wherein acceptable parameters produced deformational heating that resulted in flow stresses in the 5 – 27 MPa range. Predicted strains ranged from roughly 10 to 60 and v strain rates from 3x10^2 to 2x10^3 s^-1. The predicted parameter windows were found to be in good agreement with parameter windows reported in the literature. To aid in validating the analytical model, an experimental test bed utilizing OMC was developed to link predicted hot working conditions to the fine-grained microstructure found in FSW. Tests were performed in AA1100, AA2219, and AA6061. The experiment produced strains from ~1 to 4 and strain rates from 10^4 to 10^6 s^-1, but no significant grain refinement was observed. This implies strain, not strain rate, is the dominant factor in grain refinement, and, based on data from the literature, strains greater than 4-10 are required to achieve the grain refinement found in FSW.
Lund, Benjamin Robert, "A metal-cutting-inspired analytical framework for predicting operating parameters in friction stir welding" (2022). Dissertations. 244.