Current State of Work Presentation

1
Current State of Work Presentation

• PTP Project Mathematical and Computational
  Modelling of Friction Stir Welding

Graeme Thorn October 31st 2001
2
Structure of Presentation

• Previous modelling approaches
• Development of mathematical model
• Computational issues
• Results
• Work to come

3
Modelling Approaches (1)

• Fluid mechanics based models
• (Colegrove Bendszak et al. )
• Good for material flow in third-body region
• Does not predict material movement elsewhere
• Solid mechanics based models
• (FSW group at Univ. South Carolina )
• Good for material response far from tool
• Bad for plasticised flow in third-body region
  

4
Modelling Approaches (2)

• Heat transfer models.
• (Frigaard, Grøng, Midling ).
• Good for post weld heat analysis and for heat
  cycle calculation and microstructure
  determination.
• Does not take material movement/flow into account.

My aim is to work towards a consistent
mathematical and computational model
incorporating all features.
5
Mathematical Model Development

• Fluid-like models
• Incompressible Navier-Stokes without temperature
  dependence

6
Mathematical Model Development

• Incompressible non-Newtonian fluid with
  temperature dependence

• Good for third-body region and will be retained

7
Mathematical Model Development

• Outer elastic-plastic material Prandtl- Reuss
  plasticity

Simplest model for plasticity will be added to
and superseded to include other features of real
material deformation.
8
Computational Issues
Problem has complex geometry moving boundary
flow problem

• My approach is to use overlapping CHIMERA grid
  built using OVERTURE framework (http//www.llnl.go
  v/CASC/Overture)

Grid goes in here
9
Results

• Validation of OVERTURE solver
• Newtonian flow

10
Work To Come

• Extension of computation to non-Newtonian flow
• Move from 2D to 3D
• Increase complexity of mathematical model