I am reading "Nonlinear Electromechanics", by Dmitry Skubov and Kamil S. Khodzhaev, Springer 2008. Here is the relevant and freely available chapter.

Essentially, a loop of area $S$, mass $m$, moment of inertia $I$, self inductance $L$, carrying current $i$ and rigidly connected to a pendulum bar of length $l$ is placed in a magnetic field $B_0\sin\nu t$. The Lagrangian of the system is:

$$L=T-V=\frac{1}{2}I\dot{\theta}^2-\frac{1}{2}Li^2-B_0Si\sin{\nu t}\sin\theta-mgl(1-\cos\theta)$$ where $\theta$ is the pendulum's deviation from the vertical. Euler-Lagrange equation is $$\frac{\partial L}{\partial q}-\frac{d}{dt}(\frac{\partial L}{\partial \dot{q}})=0.$$ If we take the generalized variables to be $\theta$ and $i$ then we should get: $$-B_0Si\sin{\nu t}\cos\theta-mgl\sin\theta-I\ddot{\theta}=0$$ $$Li+B_0S\sin{\nu t}\sin\theta=0$$ However, the equations in the book (2.2.2) are different from these two. The first differs by signs, while the second is completely different. What is wrong?

=== EDIT: Added the equations in question === $$I\ddot{\theta}-B_0Si\sin{\nu t}\cos\theta+mgl\sin\theta=0$$ $$L\dot i+B_0S\sin{\nu t}\cos\theta \dot \theta+B_0S\nu\cos{\nu t}\sin\theta+Ri=0$$


1 Answer 1


Recall that magnetic energies add to the Lagrangian while electric energies subtract from the Lagrangian--this is easily proven looking at the Lorentz force in the Lagrangian formalism. That is to say, the Lagrangian should be defined as $$ \mathcal L=T-V+W_b-W_e $$ where $W_n$ are the magnetic ($b$) and electric ($e$) energies. Thus, your Lagrangian is $$ \mathcal L=\frac12I\dot\theta^2-mgl\left(1-\cos\theta\right)+\frac12Li^2+B_0S\sin\nu t\sin\theta i $$ which takes care of your sign error in the first part.

The missing terms in your second equation likely comes from your dissipation function (deduced from the existence of the $\dot\theta$ and $R$ terms) that will probably take the form $$ \Psi=\alpha i^2+\beta\dot\theta^2 $$ It's not (yet) clear to me the origins of $\alpha$ and $\beta$ here (which are easily determined via direct differentiation), but hopefully you'll see that it's not simply the Lagrangian from classical mechanics applied to a new situation, there are some extra features.

The system of equations required here are called the Lagrange-Maxwell equations (defined in Section 1.2 of your book): \begin{align} \frac{d}{dt}\frac{\partial W}{\partial i}+\frac{\partial\Phi}{\partial i}+\frac{\partial V}{\partial g}&=E\\ \frac{d}{dt}\frac{\partial T}{\partial\dot\theta}-\frac{\partial\left(T+W\right)}{\partial \theta}+\frac{\partial\left(\Pi+V\right)}{\partial \theta}&=Q \end{align} where

  • $W=\frac12Li^2$ is the self-inductance term
  • $V=\frac12\sum_{j=1}^N\int \epsilon E^2dV$ is the energy of the electric field
  • $\Phi=\frac12Ri^2$ is the thermal dissipation in the resistor
  • $T$ is the kinetic energy
  • $\Pi$ the potential energy
  • $E$ the emf
  • $Q$ the generalized forces

which is slightly different from the Lagrangian mechanics you've come to know.

In your case, $V=E=Q=0$, leaving \begin{align} \frac{d}{dt}\frac{\partial W}{\partial i}+\frac{\partial\Phi}{\partial i}&=0\\ \frac{d}{dt}\frac{\partial T}{\partial\dot\theta}-\frac{\partial\left(T+W\right)}{\partial \theta}+\frac{\partial\Pi}{\partial \theta}&=0 \end{align} Inserting the forms of $W$, $\Phi$, $T$, and $\Pi$ you have returns the form of 2.2.2 from your book (it does for me, at least).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.