Plate tectonics is the horizontal movement of several discrete and mostly-rigid surface-plate segments, with some plates being produced and recycled by the mantle as part of the global mantle overturn.
Lithosphere-scale thermal shear zones: The shear-heating induced localisation of lithosphere-scale deformation is suggested as a potentially important mechanism for breaking the lithosphere. We show that amongst the terrestrial planets, this type of shear localisation is expected to occur most readily on Earth (Crameri and Kaus 2010).
Hot mantle plumes: Hot mantle plumes can crucially weaken the lithosphere regionally. Thermal erosion at the plate bottom produces critical variation in plate thickness. Laterally mobile mantle plumes further excite horizontal mantle flow below the plate that leads to drips in thickened plate portions. All these mechanisms combined can lead to plate failure and subsequent subduction initiation (Crameri and Tackley 2016).
Self-consistent single-sided subduction: We have, for the first time, produced a mantle convection model that self-consistently reproduces the realistic single-sided sinking of the surface plates. A free surface and a weak crustal layer lead to nature-like single-sided subduction. The first allows the lithosphere to bend in a natural manner and the latter promotes decoupling between the colliding plates. Both factors promote asymmetry and single-sidedness of the subduction zone (Crameri et al., 2012, GRL).
Arcuate subduction trenches: We find that an arcuate shape is the natural form for trenches and slabs even in Cartesian geometries, which is contrasting the somewhat outdated 'Ping-Pong Ball Theory': The slabs and their trenches above are rather deformed by the subduction-induced toroidal mantle flow (Crameri and Tackley, 2014).
Subduction-polarity reversals: We show that a change in subduction polarity is a common feature of mantle convection. It can be induced by upper- and lower-plate age variations and has an important impact on the overall evolution of the convective system. It is, for example, shown to be a viable mechanism to reproduce new and separate subduction zones. We propose that such a polarity reversal has formed the South Sandwich subduction system after splitting from the South America subduction zone (Crameri and Tackley, 2014).
Slab-mantle dynamics: The spontaneous evolution of slab geometry is strongly affected by trench motion and subduction-induced mantle flow. This causes the slab to deform strongly in all spatial directions and through time. Fascinating dynamics of the slab like the 'slab tunneling' and of the surrounding mantle like the 'back-slab spiral flow' can develop (Crameri and Tackley, 2014).
Subduction shut-off: We show that subduction can shut off after a slab break off. This can be induced by a subduction-channel closure (i.e., an increase in upper-lower plate coupling) or by a spreading centre approaching the subduction trench (Crameri and Tackley, 2014 and Crameri and Tackley, 2015).