FSW simulation
Friction stir welding (FSW) is a solid material welding technique. The technique was invented at the Welding Institute (TWI) in United Kingdom since 1991.
Wiki : https://en.wikipedia.org/wiki/Friction_stir_welding
The technique was invented for the purpose of welding traditionally non-weldable material. Since during the entire process, the material is in a complete solid status, it will also help with eliminating many of the drawbacks from the traditional melt welding approaches, such as thermal shrinkage, unevenly distributed texture and microstructures and so on.
Previously many studies was focused on the experimental testing to detect and determine the welding processing parameters, which includes temperature, stress, material flow, and potentially other conditions during the process, for a further analysis of mechanical properties of the welds. The method is less efficient, as it usually requires large costs in preparing materials at millshop, nor is it accurate, due to the various experimental conditions which will introduce unstable variables into the comparison between different data.
As the requirement of this technique is growing for a couple of large industry companies, which includes many automobile manufacturers (Honda, Toyota, Fiat-Chrysler, GM), aircraft manufacturers (Boeing, Airbus, GKN), a better way of simulating and predicting processing conditions is urgently desirable for many R&D groups.
At Worcester Polytechnic Institute, an FSW simulation tool was developed to fulfill some of the requirement. The tool is developed by Yi Pan* during PhD thesis. The desktop software is designed for the purpose to facilitate the usage of the simulation program by client-end users in materials research side, who potential are not empowered in software developing perspective.
 |
| The cover image of the software |
Java components were implemented for the software interface (frames of windows). A core computing engine was implemented by Matlab program, to fulfill the duty on heavy calculation of a transient change in both temperature and stress fields simulated as the experimental conditions.
The mesh grid provides a finite element method to discretize a continuous field into various small elements. In each of these elements, a Lagrangian linear distribution was used for an estimation of field gradient between nodes.
Temperature and stress distribution can be conducted with a set of input values including: materials, dimensions of the workpiece, boundary conditions, tool size, processing parameters. The corresponding results are shown in different tags on the main interface panel.
The current version of the software is on a beta stage. More of the develop and polishing work is carrying on in the future.
*Author info:
Yi Pan, PhD in Materials Science - Computational Modelling and Software Development