Ellipsis Modelling Scott Dyksterhuis

Ellipsis Modelling

• Scott Dyksterhuis

• Dietmar Müller, Louis Moresi and Patrice Rey

• University of Sydney

Geological modeling is associated with a different set of requirements:

• Very large deformation of solid materials including

• accurate tracking of material (compositional) interfaces – melt?

• accurate tracking of history variables

• evolution of oriented structure

• etc…

• Only want a tool to fit data/concepts (geometry, thermal history, potential field)

• Easy use and implement

Ellipsis

• A particle-in-cell finite element code that solves 2D and 3D fluid mechanics problems with & w/o heat transfer, e.g.,

• Mantle convection
• Lithospheric deformation
• subduction

• Ellipsis works with a multigrid solver, which allows quite rapid solutions (but does constrain allowable mesh geometry)

• Solid-like materials can be given higher viscosities than fluid-like materials, or material properties can depend on temperature, depth, pressure, strain, strain rates, etc.

Ellipsis GUI

• Java based GUI for easy interaction with existing models or for creating new ones

• Allows for more complicated geometries – can even use drawings (based on seismic, reconstructions etc) to create models

• Handles all scaling automatically so user can do as much/little as they want

Ellipsis3D…

Interactive Inversion - Nimrod

• NIMROD, NIMROD/G manage the distribution of parameter-space studies on clusters / grids

• robust job distribution
• secure
• visual feedback / portal

• NIMROD/O contains a number of objective inversion algorithms to reduce the number of models evaluated

• Models are cached
• Multiple inversions to help avoid local minima

Initial Conditions

• Evolution of the geotherm through time…

• Rheology

• Strain rate

Initial Conditions

Lower Crustal Strength Models

• Relatively high lower crustal strength

Lower Crustal Strength Models

• Relatively low lower crustal strength

4 Layer models

• Strong upper mantle lithosphere

• Distributed

• Core complex

Mantle Rheology

Juxtaposed Craton

• Relatively strong “craton”

Conclusions

• Initial constrains extremely important

• temperature
• rheology

• Modelling not ready for applied examples?

• Accessibility of code important – get the geologist using the modelling software

A student’s perspective…

• Need to have some kind of online materials (rheology) database – Initial conditions

• A permanent web presence/meeting point (wiki site):

• Codes, manuals, (benchmark) results (corner store)
• community collaboration and “sounding ground”
• centralised location for groups to benchmark codes/results
• location for observationalists to pose “challenges”
• data