1. Subduction modeling
From 2D to 3D…

2. Subduction modeling About the plate tectonics 2D subduction model
Future works

3. Subduction modeling About the plate tectonics 2D subduction model
Future works

4. Plate tectonics

5. Plate tectonics
Transform boundaries

6. Plate tectonics
Divergent boundaries

7. Convergent boundaries
Plate tectonics
Convergent boundaries

8. Convergent boundaries
Plate tectonics
Convergent boundaries
Collision
Subduction

9. Subduction modeling About the plate tectonics 2D subduction model
Future works

10. First subduction model : 2D
Rectangle domain
Plate
Upper mantle

11. First subduction model : 2D
Lengths of the domain
Plate
660 km
100 km
Upper mantle
Lower mantle
3120 km

12. First subduction model : 2D
Boundary conditions
Plate
Upper mantle
No slip
Lower mantle

13. First subduction model : 2D
Boundary conditions
Plate
Free slip
Upper mantle
No slip
Lower mantle

14. First subduction model : 2D
Boundary conditions
Plate
Free slip
Upper mantle
Free slip
No slip
Free slip
Lower mantle

15. First subduction model : 2D
Boundary conditions
Plate
Free slip
No slip
Upper mantle
Free slip
No slip
Free slip
Lower mantle

16. Governing equations
Rayleigh-Taylor Instability: dense slab sinks into a less dense mantle
… conservation of mass, incompressible medium
… conservation of momentum
… deviatoric stress tensor
with
no energy equation

17. Rheology
Viscous mantle
Viscoplastic plate
Viscous
Plastic

18. The software Model setup : escript PDE solver : Finley
Escript Finley
Model setup : escript
PDE solver : Finley
Tracking of the plate : level set method
Visualization : Pyvisi, Gnuplot

19. The hardware UQ

20. 2D subduction results

21. Velocity field : vortex

22. Lifecycle : initiation

23. Lifecycle : flow reorganization

24. Limits of the 2D model

25. Necessity of the third dimension
Importance of the mantle flow
Real subductions have a limited lateral extent
Interactions on the lateral edges of subduction zones

26. Subduction modeling About the plate tectonics 2D subduction model
Future works

27. 3D subduction model Same governing equations Same tools used
… only a different setup

28. 3D subduction model no upper plate plate with subduction zone upper
mantle
Box size: big enough, for side walls to have little effect on subducting slab

29. The level set field use level set to define and track
the subducting plate
subducting slab
only subducting part of the plate is included in the FE domain

30. Symmetry hypothesis free slip side boundary
assume symmetry: model only half of the domain

31. Bottom boundary condition
fixed bottom
boundary
higher viscosity in LM and phase boundary hinder convection

32. Slab boundary conditions
free slip
top boundary
fixed
end
plate fixed, to isolate effects of tearing and rollback

33. Edge of the slab
extreme case:
free slip
tear zone

34. Top boundary condition
plate bottom
fixed

35. Sides boundary conditions
Freeslip on
the sides

36. 3D subduction results

37. Velocity field : toroidal flow

38. Lifecycle of the subduction

39. Initiation

40. Reorganization of the flow

42. Hinge position
Hinge position (km)
Time (M years)

43. Hinge position
Hinge position (km)
Time (M years)

44. Hinge rollback velocity
(mm/year)
Time (M years)

45. Hinge rollback velocity
(mm/year)
Initiation
Time (M years)

46. Hinge rollback velocity
(mm/year)
Reorganization of the mantle flow
Time (M years)

47. Hinge rollback velocity
(mm/year)
2.5 mm/year
Steady subduction
Time (M years)

48. Hinge from the top
20 My
40 My
60 My
80 My

49. Why arcs ?
Ping pong ball hypothesis

50. Why arcs ? Toroidal flow Inhomogeneities in the subducting plate
Tear resistance on the edges

51. Subduction modeling About the plate tectonics 2D subduction model
Future works

52. To be continued … weak zone, part of the FE domain fixed
normal case: tearing follows a pre-existing weak zone
subduction possible if slab strength > strength of the weak zone

53. First results

54. Thank you !  Hans Mülhaus Laurent Bourgouin Klaus Gottschaldt
… you for coming