Infinite-dimensional Hamiltonian description of a class of dissipative mechanical systems

Abstract

In this paper an approach is proposed to represent a class of dissipative mechanical systems by corresponding infinite-dimensional Hamiltonian systems. This approach is based upon the following structure: for any non-conservative classical mechanical system and arbitrary initial conditions, there exists a conservative system; both systems share one and only one common phase curve; and, the value of the Hamiltonian of the conservative system is, up to an additive constant, equal to the total energy of the non-conservative system on the aforementioned phase curve, the constant depending on the initial conditions. We describe in detail this relationship calling the conservative system “substitute“ conservative system. By considering the dissipative mechanical system as a special fluid in a domain D of the phase space, viz. a collection of particles in this domain, we are prompted to develop this system as an infinite-dimensional Hamiltonian system of an ideal fluid. By comparing the description of the ideal fluid in Lagrangian coordinates, we can consider the Hamiltonian and the Lagrangian as the respective integrals of the Hamiltonian and the Lagrangian of the substitute conservative system over the initial value space and define a new Poisson bracket to express the equations of motion in Hamiltonian form. The advantage of the approach is that the value of the canonical momentum density \pi is identical with that of the mechanical momentum m\dot{q} and the value of canonical coordinate q is identical with that of the coordinate of the dissipative mechanical system. Therefore we need not to decouple the Newtonian equations of motion into several one-dimensional ordinary differential equations.

Proposition

Proposition 2.1

9NM1

For any non-conservative classical mechanical system and arbitrary initial condition, there exists a conservative system; both systems sharing one and only one common phase curve; and the value of the Hamiltonian of the conservative system is equal to the sum of the total energy of the non-conservative system on the aforementioned phase curve and a constant depending on the initial condition.

Proof of Proposition 2.1

To begin, we must prove the first part of the Proposition 2.1, i.e. that a conservative system with Hamiltonian presented by Eq. (?) shares a common phase curve with the non-conservative system represented by Eq. (?). In other words, the Hamiltonian quantity presented by Eq. (?) satisfies Eq. (?) under the same initial condition. Substituting Eq. (?) into the left side of Eq. (?), we have

\begin{aligned} \frac{\partial{\hat{H}(q_i,p_i)}}{\partial {q_i}}&=\frac{\partial H(q_i,p_i)}{\partial {q_i}} -\frac{\partial{W_j(q_j)}}{\partial {q_i}} \\ \frac{\partial{\hat{H}(q_i,p_i)}}{\partial {p_i}}&=\frac{\partial H(q_i,p_i)}{\partial {p_i}} -\frac{\partial{W_j(q_j)}}{\partial {p_i}}. \end{aligned}

It must be noted that, although q_i and p_i are considered as distinct variables in Hamiltonian mechanics, we can consider q_i and p_i as dependent variables in constructing of \hat{H}. On the phase curve \gamma we have

\begin{aligned} \frac{\partial{{W_j(q_j) }}}{\partial {q_i}}&= \frac{\partial{(\int_{q_{j0}}^{q_j}\mathcal{F}_j(q_j) \mathrm{d} q_j+W_i(q_{i0}))}}{\partial {q_i}} =\mathcal{F}_i(q_i) \\ \frac{\partial{{W_j(q_j) }}}{\partial {p_i}}&=0, \end{aligned}

where \mathcal{F}_i(q_i) is equal to the damping force F_i on \gamma. Hence under the initial condition q_0, p_0, Eq. (?) is satisfied. As a result, we can state that the phase curve of Eq. (?) coincides with that of Eq. (?) subject to the initial condition; and \hat{H} represented by Eq. (?) is the Hamiltonian of the conservative system represented by Eq. (?).

Next, we must prove the second part of Proposition 2.1: the uniqueness of the common phase curve.

We assume that Eq. (?) shares two common phase curves, \gamma_1 and \gamma_2, with Eq. (?). Let z_1 be a point of \gamma_1 at the time t, z_2 a point of \gamma_2 at the time t, and g^t the Hamiltonian phase flow of Eq. (?). Suppose that a domain \Omega at t which contains only points z_1 and z_2, is not only a subset of the phase space of the non-conservative system (?) but also that of the phase space of the conservative system (?). Hence, there exists a phase flow \hat{g}^t composed of \gamma_1 and \gamma_2, that is the phase flow of Eq. (?) restricted by \Omega. According to the following form of Liouville’s theorem as stated in [Arnold1978]:

Theorem.

The phase flow of Hamilton’s equations of motion preserves volume: for any region D in the phase space we have

volume\ of\ g^tD=volume\ of\ D

where g^t is the one-parameter group of transformations of phase space

g^t:(p(0),q(0))\longmapsto:(p(t),q(t)),

that preserves the volume of \Omega. This implies that the phase flow of Eq. (?) \hat{g}^t also preserves the volume of \Omega. However, system (?) is not conservative, which conflicts with Louisville’s theorem; hence only a phase curve of Eq. (?) coincides with that of Eq. (?).

Definition

Definition 3.1

05W3

For two functionals F[\bm{\pi}(\bm{a}),\bm{q}(a)] and G[\bm{\pi}(a),\bm{q}(a)], there exists in a domain D of the phase space, a functional called the Poisson bracket of F and G

\{F,G\}[\bm{\pi}(\bm{a}),\bm{q}(\bm{a})]=\int_D \left[\frac{\delta F}{\delta \bm{q}(\bm{a}')}\cdot \frac{\delta G}{\delta \bm{\pi}(\bm{a}')}- \frac{\delta G}{\delta \bm{q}(\bm{a}')}\cdot \frac{\delta F}{\delta \bm{\pi}(\bm{a}')}\right]\mathrm{d} ^{2n} \bm{a},