1

I am trying to figure out how to derive the equations of motion of a particle experiencing a viscous force of magnitude $\mathbf{F}=-c\mathbf{v}$, being $c$ constant and $\mathbf{v}$ the velocity of the mass, using Lagrange formalism.

I'd like to have a:

  1. first, a general treatment of the viscous force for the motion of a free particle;

  2. then, specialize this treatment of the motion of a constrained particle, moving on a known line. As an example, I'm considering the helix, whose point coordinates are described by the parametric equations:

    $x=R\cos(\theta)$,
    $y=R\sin(\theta)$,
    $z=h(\theta)/2\pi$

    where $R$ and $h$ are constants.

basics
  • 8,365
natalie
  • 29
  • 2
    Dissipative forces like that derived from viscosity cannot be accounted in a "pure" Lagrangian formalism. However, there are some extensions, see the section on the extensions in here: Wikipedia Lagrangian Mechanics – Mauricio Oct 27 '22 at 09:44
  • 3
    I'm trying to reopen the question, with some edits. IMHO, a question about treating viscous force (and non-conservative force in general) iwith Lagrange equations is a good question. Many physicists seem to live in a "ideal" world, where everything is conservative, but it's not. Lagrange formalism is very useful and easy to use: treating non-conservative forces is a must to keep using it in the real world models with non ideal terms – basics Oct 28 '22 at 09:52
  • @Qmechanic imho, not a duplicate and quite bad answers there – basics Oct 28 '22 at 10:38
  • @Mauricio dissipation function can treat only non-conservative forces that are proportional to the time derivative of the generalized coordinates. It's not general enough in many cases – basics Oct 28 '22 at 10:40
  • 1
    @basics you are right, I was just pointing out how different the Lagrangian formalism is with an example of dissipative forces. – Mauricio Oct 28 '22 at 11:34
  • it's open! Yeah! – basics Oct 29 '22 at 20:17

1 Answers1

3

If you want to use Lagrangian formalism with non-conservative forces you need to evaluate their virtual work

$\delta W = Q_q \, \delta q$,

so that Lagrange equations read

$\dfrac{d}{dt}\dfrac{\partial \mathscr{L}}{\partial \dot{q}} - \dfrac{\partial \mathscr{L}}{\partial {q}} = Q_q$.

Free particle. For the motion of a free particle of mass $m$, experiencing drag $\mathbf{F}^d$, in absence of conservative forces

  • the Lagrangian function equals the kinetic energy $\mathscr{L} = T = \dfrac{1}{2} m |\mathbf{\dot{r}}|^2 = \dfrac{1}{2} m \mathbf{\dot{r}} \cdot \mathbf{\dot{r}}$, so that its derivatives appearing in Lagrange equations read

    $\dfrac{d}{dt}\dfrac{\partial \mathscr{L}}{\partial \mathbf{\dot{r}}} = \dfrac{d}{dt}\left( m \mathbf{\dot{r}} \right) = m \mathbf{\ddot{r}}$$\quad , \qquad$ $\dfrac{\partial \mathscr{L}}{\partial \mathbf{r}} = \mathbf{0}$

  • the virtual work reads $\delta W = \mathbf{F}^d \cdot \delta \mathbf{r}$

and the equations of motions read

$ m \mathbf{\ddot{r}} = \mathbf{F}^d$,

as expected.

Your problem: particle on a helix. In your problem, the particle is constrained to move on a helix, so that it has only one degree of freedom, namely $\theta(t)$.

  • The position of the particle reads $\mathbf{r}(t) = R \cos \theta(t) \mathbf{\hat{x}} + R \sin \theta(t) \mathbf{\hat{y}} + \dfrac{h \theta(t)}{2\pi} \mathbf{\hat{z}}$

  • The velocity of the particle in a Cartesian basis reads $\mathbf{v} = -R\dot\theta \sin \theta \mathbf{\hat{x}} + R\dot\theta \cos \theta \mathbf{\hat{y}} + \dfrac{h \dot\theta}{2 \pi}\mathbf{\hat{z}}$

  • The virtual displacement of the particle reads: $\delta \mathbf{r} = -R\delta \theta \sin \theta \mathbf{\hat{x}} + R\delta \theta \cos \theta \mathbf{\hat{y}} + \dfrac{h \delta \theta}{2 \pi}\mathbf{\hat{z}}$

  • Kinetic energy is

    $K = \dfrac{1}{2} m \left( R^2 + \left(\dfrac{h}{2\pi}\right)^2 \right) \dot\theta^2 $

  • Virtual work of drag is

    $\delta W = \mathbf{F}^d \cdot \delta \mathbf{r} = \underbrace{\mathbf{F}^d \cdot \left[ -R \sin \theta \mathbf{\hat{x}} + R \cos \theta \mathbf{\hat{y}} + \dfrac{h }{2 \pi}\mathbf{\hat{z}} \right]}_{Q_{\theta}} \delta \theta$

  • Lagrange equation becomes

    $ m \left( R^2 + \left(\dfrac{h}{2\pi}\right)^2 \right) \ddot \theta = \mathbf{F}^d \cdot \left( -R \sin \theta \mathbf{\hat{x}} + R \cos \theta \mathbf{\hat{y}} + \dfrac{h}{2 \pi}\mathbf{\hat{z}} \right)$

  • If you model the drag as $\mathbf{F}^d = -c \mathbf{v}$, you can easily find that

    $ m \left( R^2 + \left(\dfrac{h}{2\pi}\right)^2 \right) \ddot \theta = - c \left( R^2 + \left(\dfrac{h}{2\pi}\right)^2 \right) \dot \theta$.

    And if I made no mistake so far, last equation becomes $m \ddot \theta + c \dot \theta = 0$.

basics
  • 8,365