WHY this is true (sketch of the derivation): Hamilton's principle says the real path makes the action S=∫Ldtstationary (δS=0). Demanding that the first variation vanish for arbitrary wiggles δqi(t) that vanish at the endpoints, and integrating the q˙-term by parts, gives exactly the equation above. (Full derivation lives in 2.1.05-Deriving-the-Euler-Lagrange-equation.)
HOW to apply it — the universal 5-step procedure:
Count DOF → pick that many independent generalized coordinates qi.
Write T in terms of qi,q˙i (express every Cartesian velocity through your coords).
Write V in terms of qi.
Form L=T−V, compute ∂qi∂L and ∂q˙i∂L.
Plug into E–L, take the time derivative, simplify → equation of motion.
What does a cyclic coordinate give you? → a conserved generalized momentum.
Why don't tension/normal force appear? → ideal constraints do no work; coordinates absorb them.
Pendulum EOM? → θ¨=−(g/ℓ)sinθ.
Recall Feynman: explain to a 12-year-old
Imagine you want to predict how a swing moves. The hard way: track every push and pull, including the rope's tug. The easy way: just keep a "score" — how much moving energy it has minus how much height energy it has. Nature is lazy and always picks the path that keeps this score "balanced" over time. We wrote down a magic rule (the E–L equation) that, given the energy score in whatever measuring stick you like (angle, distance, whatever), automatically spits out exactly how the thing moves — and you never had to think about the rope at all.
The Lagrangian L equals
T−V (kinetic minus potential energy)
Number of generalized coordinates required
one per degree of freedom
Euler–Lagrange equation for coordinate qi
dtd∂q˙i∂L−∂qi∂L=0
Why constraint forces (tension, normal) don't appear in E–L
they do no work and are absorbed by the choice of independent coordinates
Kinetic energy in plane polar coordinates
T=21m(r˙2+r2θ˙2)
A cyclic (ignorable) coordinate (absent from L) implies
its conjugate momentum ∂L/∂q˙ is conserved
EOM of a simple pendulum
θ¨=−(g/ℓ)sinθ
Atwood machine acceleration
x¨=(m1−m2)g/(m1+m2)
Block on frictionless incline acceleration along slope
gsinα
Conserved quantity for central force motion
angular momentum Lz=mr2θ˙
First step of the E–L recipe
count degrees of freedom and pick that many coordinates
Inside the partial derivatives of L, how are q and q˙ treated
Dekho, Newton ke method mein har jagah forces, directions aur constraint forces (tension, normal reaction) ka hisaab rakhna padta hai — kaafi headache. Euler–Lagrange (E–L) method bolta hai: bhai, sirf energy sambhal. Tum bas kinetic energy T aur potential energy V likho, L=T−V banao, aur ek fix formula mein daal do. Bas, equation of motion nikal aati hai. Sabse pyaari baat — tension aur normal force kabhi aate hi nahi, kyunki acche coordinates choose karne se woh apne aap "absorb" ho jaate hain (kyunki woh koi work nahi karte).
Recipe yaad rakho — C-T-V-L-E: pehle Count degrees of freedom (utne hi coordinates lo), phir T likho, phir V, phir L=T−V, phir E–L equation dtd∂q˙∂L−∂q∂L=0 apply karo. Pendulum mein θ lo, Atwood mein ek x, incline pe slope ke along s, aur central force mein polar (r,θ). Polar mein dhyan rakho — T=21m(r˙2+r2θ˙2), sirf r˙2 likhna bada common mistake hai.
Ek aur jhakaas cheez: agar kisi coordinate (jaise θ central force mein) L mein explicitly aata hi nahi, toh uska momentum conserve ho jaata hai — angular momentum free mein mil gaya! Isko cyclic coordinate kehte hain. Yeh method itna systematic hai ki ek baar coordinates set kar liye, baaki sab mechanical calculation hai — isliye exam mein speed aur accuracy dono milti hai. Practice karo 3-4 systems, pattern haath mein aa jaayega.