Canonical Gibbs distribution and thermodynamics of mechanical systems with a finite number of degrees of freedom
1999, Volume 4, Number 2, pp. 44-54
Author(s):
Kozlov V. V.
Traditional derivation of Gibbs canonical distribution and the justification of thermodynamics are based on the assumption concerning an isoenergetic ergodicity of a system of n weakly interacting identical subsystems and passage to the limit $n \to\infty$. In the presented work we develop another approach to these problems assuming that n is fixed and $n \geqslant 2$. The ergodic hypothesis (which frequently is not valid due to known results of the KAM-theory) is substituted by a weaker assumption that the perturbed system does not have additional first integrals independent of the energy integral. The proof of nonintegrability of perturbed Hamiltonian systems is based on the Poincare method. Moreover, we use the natural Gibbs assumption concerning a thermodynamic equilibrium of bsystems at vanishing interaction. The general results are applied to the system of the weakly connected pendula. The averaging with respect to the Gibbs measure allows to pass from usual dynamics of mechanical systems to the classical thermodynamic model.
Citation:
Kozlov V. V., Canonical Gibbs distribution and thermodynamics of mechanical systems with a finite number of degrees of freedom, Regular and Chaotic Dynamics,
1999, Volume 4, Number 2,
pp. 44-54
✖
Мы используем cookie-файлы и сервис Яндекс.Метрики для анализа работы сайта, статистики и улучшения его работы. Продолжая использовать данный сайт, Вы соглашаетесь с условиями Пользовательского соглашения и условиями использования сервиса Яндекс.Метрика, а также выражаете своё согласие на использование cookie-файлов и на обработку своих персональных данных в соответствии с Политикой конфиденциальности. Вы можете запретить обработку cookies в настройках браузера.
We use cookies and Yandex.Metrica service to analyze the usage of our web-site and improve its performance. By continuing to use this website, you agree to the terms of the User Agreement and the terms of Yandex.Metrica service, and give your consent to the Cookies Policy and to the processing of your personal data in accordance with the Privacy Policy. You may deactivate cookies in your browser settings.