Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Euler lagrange equation, mechanics, need help

  1. Feb 11, 2009 #1
    Could somebody explain to me how lagrange multipliers works in finding extrema of constrained functions? also, what is calculus of variations and lagrangian mechanics, and can somebody explain to me what the lagrangian function is and the euler-lagrange equation. And, i read something about noethers theorem which seems very interesting. could somebody explain this theorem to me?

    I came up with this proof of the law of conservation of energy. It makes sense to me. Is it right?
    suppose f[x] is the force on an object with mass=1 in a force field that depends on position alone(is conservative).

    the definition of energy:
    E=.5(dx/dt)^2-S[f]dx (here S is my symbol of integration)

    because energy is constant the derivative of energy is zero.

    by the chain rule, .5d[(dx/dt)^2]/dt=(d2x/dt2)(dx/dt)

    the next step is making the substitution d[S[f]dx]/dt=(d[S[f]dx]/dx)(dx/dt), which is possible because the dx above the dt and the dx under the integral cancel out leaving a dt below the integral.

    By the fundamental theorem of calculus, the differentation undoes the integration of f, giving
    dividing through by dx/dt
    which is the original equality. QED

    Makes sense to me. the trickiest part for me was realizing that i could algebraicly manipulate the liebniz fraction d[S[f]dx]/dt=(d[S[f]dx]/dx)(dx/dt)=f*(dx/dt) to get rid of the integral.
    I read another proof that uses lagrangians but i didnt understand it.
    Could somebody explain what a lagrangian is to me
  2. jcsd
  3. Feb 11, 2009 #2
    Lagrange Multipliers are a way to solve constrained min/max problems. Basically the concept is that for some function f and some constraint g, there will be a min/max such that the gradients of f and g are parallel.

    In other words: [tex]\nabla f(x,y,z) = \lambda \nabla g(x,y,z)[/tex] where lambda is simply some scalar that makes f and g equal.

    I don't really know that much about variational calculus, but the Lagrangian is simply a way to explain that dynamics of a system. In mechanics it's just simply a formulation of KE + PE = E

    As for conservation of energy, you can't really prove it. It's an law that's based on experimental data. You're correct but you assumed that there is no energy lost in your derivation.
  4. Feb 11, 2009 #3


    User Avatar
    Homework Helper

    In a certain sense, sure, but one can show in the Lagrangian-based theory of classical mechanics that invariance with respect to time implies energy is conserved (or, perhaps more acurately, that there is a conserved quantity which we identity with the energy). So in that sense one can prove conservation of energy - to connect it to your claim, though, one must of course then prove that the world is indeed described by Lagrangian mechanics, which requires the experiments. =P
  5. Feb 23, 2009 #4
    The Lagrangian function in dynamics is usually written as
    L(q,qdot) = T(q,qdot) + V(q)
    L = Lagrangian function
    T = kinetic energy function for the system
    V = potential energy function for the system

    The calculus of variations is used to express Hamilton's Principle which is the physical principle from which the system equations are derived using the calculus of variations. Lagrange's equations simply short cut the process, going straight to the equations that result from the application of the calculus of variations to Hamilton's Principle. In most cases, this works just fine and nothing is lost. The equations of motion are developed in a very quick, efficient manner.

    In a few circumstances, particularly in dealing with mixed systems (such as electromechanical systems, electroacoustic system, mechanical-hydraulic systems, discrete-continuous systems, etc), there are so-called "natural boundary conditions" that apply at the interface between the two systems which are not evident when Lagrange's equations are used but which come out naturally through the direct application of the calculus of variations to Hamilton's Principle.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook