The Lagrangian and the Condition for Stationary Action

The initial and final positions and times are fixed, and a single number called the action is calculated for each path.

時間幅 3π/2 s。どちらの経路の作用が小さいか?

The equations explain the phenomenon

  1. 1Constructing the LagrangianL=⁠T − U=⁠½ m x² − U(x)……
  2. 2Integrating L along the path gives the action S.S=⁠∫t₁t₂ L(x, x) dt……
  3. 3A ball thrown vertically upward illustrates how varying the path changes the action.x(t)→⁠x(t) + ε η(t)⁠,η(t₁) = η(t₂) = 0……
  4. 4Expanding L gives the 1st-order change in action.δS=⁠ε ∫t₁t₂(∂L/∂x · η+ ∂L∂x · η)dt……
  5. The aim is to factor out η inside the integral to obtain a condition that holds for every η. The term containing η prevents this.
  6. 5Integrating the term containing η by parts allows η to be factored out.δS=⁠ε ∫t₁t₂[∂L/∂x− ddt(∂L/∂x)]η dt……
  7. 6The condition for δS = 0 for every variation follows.ddt (∂L∂x)=⁠∂L∂x……
  8. 7The stationarity condition yields Newton's equation of motion.m x=⁠−dUdx……
  9. 8Calculating the 2nd-order change in action for a springS(A) − S₀=⁠(m A² τ4)[(πτ)² − km]……
  10. 9Describing pendulum motion using an angleL=⁠½ m ℓ² θ² + m g ℓ cosθ→⁠θ = −(gℓ) sinθ……
  11. Further applications to other types of motion
  12. 10The method also applies to a double pendulum.L(θ₁, θ₂, θ₁, θ₂)→⁠2 equations……
  13. 11Rewriting the description using the HamiltonianH = p x − L⁠,x = ∂H∂p,p = −∂H∂x……
  14. 12Plotting motion in the plane of angle and generalized momentumH(θ, p)=⁠Constant……