Numerical modeling and validation for the development of tool geometry and material for friction stir welding of thick copper

Abstract

This thesis catalogs the development of a numerical model for the Friction Stir Welding (FSW) of large sealed copper canisters used in the disposal of nuclear waste. FSW is a highly complex process, with many interconnected physical phenomena that determine the quality and effectiveness of the weld. FSW of copper is particularly challenging due to the increased processing temperature required. Specifically, for the purpose of FSW 50 mm thick copper canisters, Posiva Oy, SKB, and TWI jointly developed the current process. While the current process produces acceptable welds, it was noted that further improvements could be made to the probe material and geometry. These improvements were not explored during the original development cycle due to the high cost associated with the manufacturing and testing of full-scale probe prototypes. The numerical model developed in this thesis combats this problem by allowing multiple probe prototypes to be simulated, and their performance compared to the original probe design.

The model developed in this work uses a highly refined mesh near the probe surfaces, giving it the capability to simulate very fine details in the probe geometries. A rigid sliding mesh technique gives the model the ability to simulate complex 3D material flow through time. A strain rate and temperature dependent material model, based on the Sheppard-Wright flow stress equation, was also created specifically for copper. The model was validated by comparing the simulation’s results to evidence seen on actual probes. The probes simulated in this work were previously created using analytical design criteria and were focused on improving material flow properties, especially near the probe tip. The results of the simulations give proof of concept that comparative analysis of different probe designs can lead to a more optimized probe without the need for prototyping of every design revision.

The current probe material, Nimonic 105, was chosen for its high temperature strength and stability. Other material options could exist but need to be systematically evaluated. The Low Cycle Fatigue (LCF) test protocol developed in this work is designed to evaluate a material’s performance in a FSW environment. Once a baseline for Nimonic 105 has been established using this protocol, other materials could be evaluated using this protocol.