Computational Thermodynamics

In the blog of Prof Claes Johnson

I have found a link to his book Computational Thermodynamics that immediately caught my attention as I used to work in this area some time ago. Unfortunately his book happens to have nothing with what people understand under Computational Thermodynamics, rather his goal was to suggest a new Second Law.

The book makes pretty strange impression. I will make just two citations.

In this book we present a constructive deterministic fluid mechanics model of the thermodynamics of a slightly viscous gas/fluid based on finite precision computation by a weighted least squares stabilized finite element method for the Euler equations with slip boundary conditions at solid boundaries, referred to as Euler General Galerkin or EG2. For definiteness we consider a perfect gas/fluid, but extension to more general state equations for the pressure is possible.

Classical thermodynamics is based on a 1st Law stating conservation of total energy combined with a 2nd Law involving entropy in the form
             TdS = dE + pdV, dS >= 0,          (6.1)
where T is (absolute) temperature, S entropy, E internal energy, p pressure and V volume, and d represents change. The enigma of classical thermodynamics is to give the entropy S a physical meaning and to motivate why dS >= 0.

The second statement is just wrong. In Eq (6.1) nobody requires dS >= 0. I would advise Prof Johnson to talk with his neighbor, Prof Bo Sundman

who is working in the area of Computational Thermodynamics for quite awhile. I guess this would help him first to learn classical thermodynamics properly, second to try to use his ideas on something real.

Comments are closed.