On the Grand Challenges in Physical Petrology: the Multiphase Crossroads

Thursday, 18 December 2014: 2:10 PM
George W Bergantz, Univ Washington, Seattle, WA, United States
Rapid progress in experimental, micro-analytical and textural analysis at the crystal scale has produced an unprecedented record of magmatic processes. However an obstacle to further progress is the lack of understanding of how mass, energy and momentum flux associated with crystal-rich, open-system events produces identifiable outcomes. Hence developing a physically-based understanding of magmatic systems linking micro-scale petrological observations with a physical template operating at the macro-scale presents a so-called “Grand Challenge.”

The essence of this challenge is that magmatic systems have characteristic length and feedback scales between those accessible by classical continuum and discrete methods. It has become increasingly obvious that the old-school continuum methods have limited resolution and power of explanation for multiphase (real) magma dynamics. This is, in part, because in crystal-rich systems the deformation is non-affine, and so the concept of constitutive behavior is less applicable and likely not even relevant, especially if one is interested in the emergent character of micro-scale processes. One expression of this is the cottage industry of proposing viscosity laws for magmas, which serves as “blunt force” de facto corrections for what is intrinsically multiphase behavior. Even in more fluid-rich systems many of these laws are not suitable for use in the very transport theories they aim to support.

The alternative approach is the discrete method, where multiphase interactions are explicitly resolved. This is a daunting prospect given the numbers of crystals in magmas. But perhaps all crystals don’t need to be modeled. I will demonstrate how discrete methods can recover critical state behavior, resolve crystal migration, the onset of visco-elastic behavior such as melt-present shear bands which sets the large-scale mixing volumes, some of the general morpho-dynamics that underlies purported rheological models, and transient controls on the emergence and dissipation of distinct thermodynamic states. As simulations with 106 – 107 crystals are now possible both the local, micro-scale crystal processes as well as the larger scale processes controlled by particle-particle-fluid interactions, can be simultaneously resolved.