In their fine Reference Frame “Microscopic Irreversibility and Chaos” (Physics Today, August 2006, page 8), Jerry Gollub and David Pine remark that they don't know why there is a threshold for irreversibility in the Taylor fluid rotation experiment. They also ask whether the origins of microscopic irreversibility can be usefully explored for other areas of physics. Recent experiments on the composition dependence of reversibility in glass transitions have uncovered surprising new effects that are relevant here.
It is often assumed that glasses have not crystallized because they are mixtures of different compounds that have not been able to phase separate during quenching—in other words, that glass formation is primarily a kinetic phenomenon related to chemical chaos. However, several compounds—silica (SiO2) and arsenic sulfide (As2S3), for example—are good glass formers even when pure. Similarly, it has long been supposed that all glass transitions are irreversible, with the degree of irreversibility varying slowly with composition. Using phase-modulated calorimetry, Punit Boolchand and colleagues found that network glass alloys show a reversibility window, with abrupt edges (thresholds). 1 Outside the window the degree of irreversibility does vary slowly with composition, but within the window the irreversibility is smaller than outside by a factor of 10.
Theoretical models at present are primitive and merely relate the reversibility window not to dynamics and boundary conditions, as in hydrodynamics, but to statics and space filling. It may be that a better understanding of reversibility will be found not in fluids but in glasses, with a theory that includes hydrodynamic concepts.
