4.6.20 · Maths › Ordinary Differential Equations
Jab aap spherical symmetry wale physics problems solve karte ho (kisi planet ki gravity, charge ka electric potential, ball par heat), toh equation ka angular part hamesha ek specific ODE mein convert ho jaata hai. Woh ODE hai Legendre's equation . Iske "sundar" solutions — polynomials jo poles par blow up nahi karti — woh hain Legendre polynomials P n ( x ) . Toh sirf is ek equation ko master karo aur poori 3D physics ki family unlock ho jaati hai.
Definition Legendre's differential equation
Ek parameter n ke liye (jo usually non-negative integer hota hai), Legendre's equation hai
( 1 − x 2 ) y ′′ − 2 x y ′ + n ( n + 1 ) y = 0 , x ∈ ( − 1 , 1 ) .
Equivalently, self-adjoint form mein:
d x d [ ( 1 − x 2 ) y ′ ] + n ( n + 1 ) y = 0.
[ − 1 , 1 ] par bounded polynomial solution hai Legendre polynomial P n ( x ) .
x = cos θ KYU aata hai? Spherical coordinates mein angular equation mein θ variable hota hai; x = cos θ substitute karne par woh upar wali form mein aa jaata hai, jahan x ∈ [ − 1 , 1 ] . Endpoints x = ± 1 sphere ke poles hain — aur woh ODE ke singular points hain.
Intuition Dono forms ek jaisi kyun hain
Self-adjoint form expand karo:
d x d [ ( 1 − x 2 ) y ′ ] = ( 1 − x 2 ) y ′′ + d x d ( 1 − x 2 ) ⋅ y ′ = ( 1 − x 2 ) y ′′ − 2 x y ′ .
n ( n + 1 ) y add karo aur standard form wapis mil jaati hai. Yeh kyun important hai: self-adjoint (Sturm–Liouville) shape guarantee karta hai ki solutions orthogonal hain — woh gold property jo hum har jagah use karte hain.
Series KYU? x = 0 ek ordinary point hai (y ′′ ka coefficient 1 − x 2 wahan nonzero hai), toh ek plain Taylor series y = ∑ a m x m converge karti hai aur kaam karti hai.
Assume karo
y = ∑ m = 0 ∞ a m x m .
Toh y ′ = ∑ m a m x m − 1 aur y ′′ = ∑ m ( m − 1 ) a m x m − 2 . Plug in karo:
( 1 − x 2 ) ∑ m ( m − 1 ) a m x m − 2 − 2 x ∑ m a m x m − 1 + n ( n + 1 ) ∑ a m x m = 0.
Yeh step kyun? Hum har sum ko same power x m par re-index karte hain:
∑ m ( m − 1 ) a m x m − 2 → ∑ ( m + 2 ) ( m + 1 ) a m + 2 x m (shift m → m + 2 ).
Baaki already x m carry kar rahe hain.
x m ka coefficient collect karo aur 0 set karo:
( m + 2 ) ( m + 1 ) a m + 2 − m ( m − 1 ) a m − 2 m a m + n ( n + 1 ) a m = 0.
a m terms group karo: − m ( m − 1 ) − 2 m + n ( n + 1 ) = n ( n + 1 ) − m ( m + 1 ) . Toh
n ka jaadu — termination
Numerator m ( m + 1 ) − n ( n + 1 ) dekho. Jab m = n hota hai, yeh 0 ban jaata hai, toh a n + 2 = 0 , aur phir us parity ke saare aage ke coefficients khatam ho jaate hain. Infinite series collapse hokar degree n ka polynomial ban jaata hai. Yahi reason hai ki n ka non-negative integer hona ek clean polynomial P n deta hai. Non-integer n ke liye series kabhi terminate nahi karti aur x = ± 1 par blow up ho jaati hai.
Hum P n ( 1 ) = 1 convention se normalize karte hain.
n = 0
Equation: ( 1 − x 2 ) y ′′ − 2 x y ′ = 0 . Ek constant kaam karta hai: y = const. P 0 ( 1 ) = 1 ke saath:
P 0 ( x ) = 1.
Yeh step kyun? n = 0 ke liye recurrence a 0 ke aage sab kuch khatam kar deta hai; constant hi polynomial hai.
n = 1
Odd series lo (a 0 = 0 ). n = 1 ke saath recurrence: a 3 = 6 1 ⋅ 2 − 1 ⋅ 2 a 1 = 0 . Toh y = a 1 x . Normalize P 1 ( 1 ) = 1 ⇒ a 1 = 1 :
P 1 ( x ) = x .
n = 2
Even series. a 2 = 2 0 − 2 ⋅ 3 a 0 = − 3 a 0 ; a 4 = 12 2 ⋅ 3 − 6 a 2 = 0 . Toh y = a 0 ( 1 − 3 x 2 ) .
Normalize: x = 1 par, a 0 ( 1 − 3 ) = − 2 a 0 = 1 ⇒ a 0 = − 2 1 .
P 2 ( x ) = 2 1 ( 3 x 2 − 1 ) .
Yeh step kyun? Normalization P n ( 1 ) = 1 ek free constant ko pin karta hai.
P n ( − x ) = ( − 1 ) n P n ( x ) — even n even function, odd n odd.
P n ( 1 ) = 1 , P n ( − 1 ) = ( − 1 ) n .
P n ki degree exactly n hai; leading coefficient 2 n ( n ! ) 2 ( 2 n )! .
Common mistake Common errors ko steel-man karo
Mistake 1: "Legendre's equation ke dono solutions polynomials hain."
Kyun sahi lagta hai: tumne ek polynomial P n dhundha, toh symmetry se lagta hai doosra bhi hoga. Sach yeh hai: yeh 2nd-order ODE hai, toh ek doosra , independent solution Q n ( x ) hota hai — Legendre function of the second kind — jo woh series hai jo terminate nahi karti aur x = ± 1 par logarithmically blow up ho jaati hai . Hum ise discard karte hain kyunki yeh poles par unbounded hai.
Mistake 2: x = 0 ordinary hai yeh dekhke x = ± 1 ke singular hone ko bhool jaana.
Fix: equation interior mein theek hai lekin ( 1 − x 2 ) ± 1 par vanish hota hai, jo unhe regular singular points banata hai. Poles par bounded hona physical selection rule hai jo force karta hai n ∈ { 0 , 1 , 2 , … } .
Mistake 3: Recurrence numerator ko m ( m − 1 ) − n ( n + 1 ) padhna.
Kyun sahi lagta hai: y ′′ term akela m ( m − 1 ) deta hai. Fix: tumhe − 2 x y ′ term se − 2 m bhi add karna hai, jo m ( m + 1 ) − n ( n + 1 ) deta hai. Isse bhool jaana termination tod deta hai.
Recall Forecast-then-verify
Q: Aage padhne se pehle predict karo P 4 ( 1 ) kya hoga aur P 4 even hai ya odd.
A: P 4 ( 1 ) = 1 (hamesha, normalization se); P 4 even hai kyunki n = 4 hai. (P 4 ( x ) = 8 1 ( 35 x 4 − 30 x 2 + 3 ) .)
Recall Feynman: 12-saal ke bachche ko samjhao
Socho tum ek beach ball par temperature ko sirf smooth "shape patterns" use karke describe kar rahe ho. Ek flat pattern hai (har jagah same), ek top-vs-bottom pattern hai, ek "ring" pattern hai, aur aise hi aur. Legendre polynomials exactly woh building-block patterns hain. Inhe dhundhne ka rule ek clean equation hai, aur cool trick yeh hai: sirf woh patterns allowed hain jo ball ke dono poles par crazy nahi jaate . Inhe sahi matra mein add karo aur tum ball par koi bhi smooth temperature describe kar sakte ho.
"Self-Adjoint Sphere Selects Sane Solutions." Self-adjoint form → orthogonality; sphere → x = cos θ ; selects → termination se integer n ; sane → bounded P n (wild Q n ko bahar phenko).
Pehle chaar: "1, x, half(3x²−1), half(5x³−3x)" — "one, ex, half three-ex-squared-minus-one, half five-ex-cubed-minus-three-ex."
Legendre's equation (standard form) ( 1 − x 2 ) y ′′ − 2 x y ′ + n ( n + 1 ) y = 0
Legendre's equation self-adjoint form mein d x d [( 1 − x 2 ) y ′ ] + n ( n + 1 ) y = 0
Series coefficients ke liye recurrence relation a m + 2 = ( m + 2 ) ( m + 1 ) m ( m + 1 ) − n ( n + 1 ) a m
Integer n ke liye series terminate kyun hoti hai? m = n par numerator m ( m + 1 ) − n ( n + 1 ) = 0 ho jaata hai, toh a n + 2 = 0 aur saare aage ke same-parity coefficients khatam, degree-n polynomial milti hai.
P 0 , P 1 , P 2 , P 3 1 , x , 2 1 ( 3 x 2 − 1 ) , 2 1 ( 5 x 3 − 3 x )
Rodrigues' formula P n ( x ) = 2 n n ! 1 d x n d n ( x 2 − 1 ) n
Orthogonality relation ∫ − 1 1 P m P n d x = 2 n + 1 2 δ mn
Normalization convention P n ( 1 ) = 1 (aur P n ( − 1 ) = ( − 1 ) n )
P n ki parityP n ( − x ) = ( − 1 ) n P n ( x )
Doosra solution kya hai aur ise discard kyun karte hain? Q n ( x ) , Legendre function of the second kind; yeh x = ± 1 par unbounded hai (log singularity), toh physically reject kiya jaata hai.
x = 0 ordinary point kyun hai lekin x = ± 1 singular?y ′′ ka coefficient ( 1 − x 2 ) 0 par nonzero hai lekin ± 1 par vanish karta hai (regular singular points = sphere ke poles).
Power Series / Frobenius Method — jaise humne y = ∑ a m x m banaya.
Sturm-Liouville Theory — self-adjoint form ⇒ orthogonality aur real eigenvalues.
Laplace's Equation in Spherical Coordinates — jahan x = cos θ aur Legendre's equation paida hoti hai.
Fourier Series — same idea, orthogonal basis expansion, sines ki jagah polynomials.
Associated Legendre Functions & Spherical Harmonics — agla step (m = 0 angular modes).
Generating Functions — 1 − 2 x t + t 2 1 = ∑ P n ( x ) t n multipole expansion se connect karta hai.
Spherical symmetry physics
Angular equation in theta
Power series y = sum a_m x^m