Except for particular HPT straining way, the minimum thickness of fragments separated by HABs is generally some MISTIC invariant (1000 b i.e. about 200 nm) regardless of deformation scheme.
Apparently this suggests "penetration" of deformed matter through a fixed totality of interfaces. In other words, permanent local migration of boundaries with their repeated annihilation necessary to maintain the fragment size despite of deformation. Which is this kind of
LOW TEMPERATURE migration, nobody could explain me.
Probably, it is similar to very local GB migration in the so called "continuous REX" by Humphris, but it is not obvious as well...
Yes, there is a kind of near-stabilization of the dislocation boundary spacing at high strains that can only be due to some form of boundary migration and annihilation to compensate the boundary creation. Some recent work by the Riso group on heavily deformed aluminium suggests boundary elimination at relatively low temperatures by Y junction migration; see for example their recent paper in Acta Materialia - "In situ observation of triple junction motion during recovery of heavily deformed aluminum " by Tianbo Yu et al (Acta March 2015)
Thanks a lot for the reference! By the title, it seems similar to the Humphrey's mechanism indeed. Anyway, it (migration) is still amazing from the view of GB mobility at such temperatures.
BTW, for Al (particularly) room temperature is in no way low, sooner WARM.
Dear Prof. Driver!
the recommended papers by Riso consider Al at rather high Homologous temperatures 0.4 to 0.5 , whereas the Y-junction migration of interest takes some hundreds sec.
Meanwhile the processes we mean go synchronically with deformation usually taking
tens sec or less. Moreover there a lot of metals fragmented under homologous temp lower than 0.3. Enough to mention Molibdenuium (!!) thoroughly studied in USSR in
1980th. In all, the Riso scenario is hardly relevant to the problem. Probably we should think of transversal migration through low-angle bridges (like in Humphrey's cont REX) rather than along thin fragments. But clear model of that is still waited...
So we have only problem as yet and almost no hints how to solve this.
- Badges
- Science method