Paper detail

Reply to W. G. Hoover [arXiv:1204.0312v2]

In response to W. G. Hoover's comment [arXiv:1204.0312v2] on our work [arXiv:1203.5968], we show explicitly that the divergence of the velocity field associated with the Nosé-Hoover equations is nonzero, implying that those equations are not volume preserving, and hence, as often stated in the literature, are not Hamiltonian. We further elucidate that the trajectories {q(t)} generated by the Nosé-Hoover equations are generally not identical to those generated by Dettmann's Hamiltonian. Dettmann's Hamiltonian produces the same trajectories as the Nosé-Hoover equations only on a specific energy shell, but not on the neighboring ones. This fact explains why the Nosé-Hoover equations are not volume preserving. The Hamiltonian that we put forward with [arXiv:1203.5968] instead produces thermostated dynamics irrespective of the energy value. The main advantage of our Hamiltonian thermostat over previous ones is that it contains kinetic energy terms that are of standard form with coordinate-independent masses and consequently is readily matched in laboratory experiments.