Customer Reviews

Statistical Mechanics, 2nd Edition

5 star:		(4)
4 star:		(5)
3 star:		(0)
2 star:		(2)
1 star:		(9)

 

 

I learned statistical mechanics from `Statistical Mechanics' by K. Huang and `Statistical Mechanics' by S.K. Ma. In my opinion, most books on the elementray principles, including the book by K. Huang, are too complicated for beginners. The best treatments as far as I know are given by `Statistical Physics, part I' by Landau and Lifshize and the one by S.K. Ma. Another weak points in Huang's book are the discussions about ideal quantum gases. It presents formal mathematical methods to study this problem. I think that this approach is also too complicated for beginners and sometimes bury the relevant physical ideas. In this part, the best treatment for the ideal Bose gas is given by `Statistical mechanics' by T.D. Lee (in Chinese), and for the ideal Fermi gas given by S.K. Ma.

For these parts, I should give 3 stars. However, the strong parts of Huang's book is the chapters on the advanced topics. The writting is compact and clear. They can be served as a good introduction to the modern theory of critical phenomena and superfluidity. Further, they are useful references for research. In addition, the formal manipulation for quantum ideal gases is necessary for research though it seems a little bit complicated for students. For all these, I gave 4 stars to this book.

Finally, I should say that the approach of Huang's book is not based on the kinetic theory though it spends a few chapters on this aspect. The reason why the kinetic theory is put before the chapters on SM, in my opinion, is to emphasize the important role played by collisions between particles to establish thermal equilibrium and the validity of the basic assumption of SM, as indicated by S.K. Ma in his book. I think previous reviews about this are misleading.

I just completed a class taught from Huang. It was rocky. If this is your first time through a serious stat mech class, you have difficulty with thermodynamics, or you are the sort of person that learns by doing, then this probably isn't the book for you.

That being said, if you have a strong background in stat mech and thermo, you'll probably find Huang quite enlightening. Huang's approach isn't the most pedagogically sound, but he is more insightful than most. (In the sense of what he presents, not necessarily how he presents it.)

Additionally, while there are errors in the book, there aren't so many that it makes it unreadable -- the book is hard to read based on its own construction.

I managed to do well in the course, but it was by constantly referencing other books. If this is your first time through, I recommend reading the material of the current chapter in another book first, running a couple of example problems from that book, and then reading through Huang. It seemed to work better as a reference into usually uncovered topics than as a textbook. The first few chapters on thermodynamics and (non-quantum) statistical mechanics are probably best learned from another source.

It might or might not be a good idea to use this book as a textbook for a course of statistical mechanics. However, it contains some material which is difficult to find elsewhere. First of all, there is a very nice exposition of Lars Onsager's own solution of the 2D Ising model which I found easy to follow. The book also contains a nice illustration of renormalization ideas with the 1D Ising model. The first edition of the book appeared in 1960-es and contained some new pedagogical ideas which appealed to many physicists. The book was almost immediately translated into Russian which emphasizes its significance.

Few books on Statistical Mechanics present a treatment with a grounding in the Boltzmann Transport Equation. It is more usual the case that a statistical approach is adopted, in which the canonical (or other) ensemble is arrived, and subsequent results somehow refer back to the the ensembles. Huang seems to wish to proceed from a strongly microscopic and kinetic point of view. There must be great strength in his approach, as statistical phenomena, in reality, is built up from this semi-classical sort of physics. If you like kinetic theory, then buy it. However, for those of us who find more comfort in relating to a more statistical approach, this book is fearsomely unreadable.

But do read it for a description of the Boltzmann Transport equation.

I would suggest that the previous reviewer get him/herself a tutor. (The formula Huang uses, log n! = n log n - 1, is correct for large numbers.) Statisticam Mechanics is a notoriously hard subject to present lucidly and clearly. Huang does a fairly good job. I was a little disappointed with his treatment of critical phenomena at the end of the book; for a much better and more comprehensive treatment, read Nigel Goldenfeld's "Lectures on phase transitions and the renormalization group".

As repeated by reviewers below, this is NOT STANDARD textbook on statistical mechanics because it stresses the kinetic theory. If reader would like to learn the equilibrium theory, this book might be embarassing. However, this book is recommended to anyone who is interested in "unusual" viewpoint. I prefer Huang to other numerous too standard textbooks. But I am a little disappointed that some interesting topics are removed in the new edition, such as the Chapman-Enskog method.

Now, it is NOT SO BAD.

a generation of physicists learnt statistical physics from this book.
It contains many insightful, yet terse remarks on the fundamental
principles of statistical mechanics. Some contents are hard to
find in the other textbooks for graduate students.

In my 18 years of formal education, this is the worst textbook I have ever used in any subject. His writing style is awful, the book is littered with typos, his notation is completely different from any I've ever seen, and attempting the homework problems is similar to trying to write an essay in Korean with only a travel guidebook as a reference. Even the paper and cover is of substandard quality. If you are an instructor and are reading this, for the love of god, DON'T USE THIS FOR YOUR COURSE.

Huang approaches the subject as a series of proofs: he does not make physical arguments, and his writing is wooden. Instructors--avoid this book!

Some have said that this book approaches stat mech from the refreshing view of kinetic theory. But it leaves out the Fokker-Planck and Langevin approaches, by which the Boltzmann equation is usually solved. Anyone interested in this approach would be *far* more rewarded by Landau's Physical Kinetics.

Anyone interested in Gibbs theory should consult Landau or Sommerfeld.

Anyone who wants good problems (and real applications) would be better served by the canonical McQuarrie.

Anyone who wants a feel for what the subject *actually now is* should see Kadanoff or Chandler. Actually I think allowing students to leave stat mech without seeing the monte carlo algorithm or solving a stochastic equation is a crime.

The reviewer below who said that this book pursues primarily a kinetic theory - Boltzmann Transport Equation approach, got it right. It really is a fearsome, and by and large, pointless read. Our professor used this book in our stat. mech. class back in 1992. He also used Mahan's Many Particle Physics book in our solid state course and de Genne's Superconductivity text in our superconductor course, so that gives you an idea of what kind of person likes Huang. Most students I've talked to feel that this text is the worst sort of student pain. The pain you feel when after exerting colossal effort trying to understand, you realize at the end of the semester that you didn't learn anything, and that you could have, if only the instructor had chosen one of any number of better books. I am completely mystified as to why and how this book has reached a 3rd edition. Perhaps there are too many physics professors out there who don't care about pedagogy.