The doctrines of grace as seen by a nuclear physicist

Thursday, November 16, 2017

Entropy

Today I start lecturing on entropy in my thermo class (which begins in 5 minutes). It is my favorite topic in all of physics. I am felling quite giddy!

By the way, we no longer teach entropy in terms of order and disorder. At least that is not the primary paradigm (is that redundant?), as it once was. That's now considered sooooooo 20th century.

In a macro sense it is taught as dS = dQ/T, as always. In a microscopic sense we use S = k ln(W), where W = the number of configurations (microstates) for a give macroscopic state (again, not new). But when we relate higher entropy to many things, the things tend to be more configurations, missing information, and probability probability rather than disorder. The latter may me touched upon (I do) but it is too nebulous to say that the second law implies we move from order to disorder. It is much more precise (if not as sexy) to say we move to macrostates with more underlying micro states. The really cool part is connecting the 19th century equation dS = dQ/T. developed before atomic theory, to the Boltzmann equation, S = k ln(W) which is inherently microscopic.

curious (as a product of the 20th century :) ) -- in what terms is entropy taught these days?

ReplyDeleteIn a macro sense it is taught as dS = dQ/T, as always. In a microscopic sense we use S = k ln(W), where W = the number of configurations (microstates) for a give macroscopic state (again, not new). But when we relate higher entropy to many things, the things tend to be more configurations, missing information, and probability probability rather than disorder. The latter may me touched upon (I do) but it is too nebulous to say that the second law implies we move from order to disorder. It is much more precise (if not as sexy) to say we move to macrostates with more underlying micro states. The really cool part is connecting the 19th century equation dS = dQ/T. developed before atomic theory, to the Boltzmann equation, S = k ln(W) which is inherently microscopic.

ReplyDelete