Trends in Welding Home      Event Overview      ASM Homepage
Back to "Session 18 - Friction Stir Welding - Modeling" Search
    Back to Main Search

Wednesday, June 4, 2008 - 3:40 PM

A predictive thermo-fluid modelling approach for computing temperature cycles and metallurgical change in FSW

B. Ducoeur, B. de Meester, Université catholique de Louvain, Louvain la Neuve, Belgium; T. De Vuyst, L. D'Alvise, CENAERO, Gosselies, Belgium

The correct computation of the temperature and flow fields is a necessary step towards the prediction of the metallurgical changes that occur in a Friction Stir Welded joint. This is also required for the correct prediction of the residual stresses distribution. Currently most numerical models rely on the measurements of the heat flux to be used as input data into the numerical model. This means that these models are not predictive.
In this paper a three-dimensional thermo-fluid model is used to predict the temperature and flow fields around a FSW tool. It consists of a sequential flow and thermal calculation. The flow problem is solved in an Eulerian formulation with a mixed velocity-pressure discretisation. The plastic work is then calculated from flow simulation and is used as a heat source for the local thermal calculation. Experimental input data relative to the heat sources are therefore no longer needed making the model predictive.  Those features are included in the MORFEO finite element code developed at CENAERO that is used for all simulations.
Simulation results concerning the resulting temperature field in the workpiece and the material flow are presented.  They are compared with experimental data: thermocouple measurements in the welded joint, backing plate and tool and material flow visualization experiments. The experiments were performed at UCL on AA7449 T79 sheets.  The model is validated for a range of operating parameters (welding speed, tool rotation speed,...)
Also, the results of the thermo-fluid model are post-processed to extract temperature, strain-rate and strain histories in a section across the weld. This data is used as input data for Myhr and Grong type metallurgy model. The calculated results are shown to be in good agreement with the hardness distribution measured on sections transverse to the welds.

Summary: A predictive 3D thermo-fluid model is developed to compute the temperature and flow fields around a FSW tool. The results of this model are then post-processed to apply a Myhr and Grong type metallurgy model to describe the microstructural and hardness variation across a transverse section of AA7449T79 welded joints.