The talk is largely based on a recent paper with Nandan Pakhira.

Here is one of the slides I am working on.

Some of the points I want to make here are the following.

There is experimental data on real systems.

This graph shows how for liquid 3He the shear viscosity varies by 3 orders of magnitude.

Here the shear viscosity decreases with increasing temperature (while the mean-free path gets shorter). This is counter-intuitive, a feature shared by dilute classical gases.

Fermi liquid behaviour is seen at low temperatures with the viscosity scaling with the scattering time (and mean free path) which is inversely proportional to T^2.

The last point is the most important.

At "high" temperatures, above the Fermi liquid coherence temperature (about 50 mK), the shear viscosity becomes comparable to the value n hbar (where n is the fluid density), which was conjectured by Eyring 80 years ago to be

**a minimum possible value.**

This also corresponds to the Mott-Ioffe-Regel limit (where the mean free path is comparable to the inter particle spacing) (bad metal).

Aside: I still think it is amazing that when scaled with the density the viscosity [a macroscopic quantity] has the same units as Planck's constant. Just like h/e^2 is in ohms.

An earlier post considers similar issues for the unitary Fermi gas, that can be realised in ultra cold atom systems.

The

**big question**addressed by the paper and talk is whether the comparable values are just a coincidence and whether by increasing the interactions you can go below the quantum limit.

## No comments:

## Post a Comment