Classification — elliptic, parabolic, hyperbolic (discriminant test): Foundations

4.7.1 · D1 · Maths › Partial Differential Equations › Classification — elliptic, parabolic, hyperbolic (discrimina

Isse pehle ki tum us ek number par trust kar sako, tumhe har us symbol mein fluent hona hoga jisse woh bana hai. Yeh page kuch bhi assume nahi karta. Hum har piece banate hain, use ek picture se anchor karte hain, aur exactly batate hain ki topic ko yeh kyun chahiye.

0. Do variables ka function kya hota hai?

Yahan sab kuch ek flat sheet par hota hai. Us sheet par ek point choose karo do coordinates se: ek horizontal position $x$ aur ek vertical position $y$ . Har aisi jagah par ek single number $u$ baitta hai.

Figure s01 — landscape. Ek 3-D surface flat $x y$ -floor ke upar baithi hai: har floor-point $(x, y)$ ko height $u (x, y)$ tak lift kiya jaata hai. Yeh woh picture hai jo tumhare dimaag mein baaki poore page ke liye rehni chahiye — baad mein har symbol is surface ki shape ke baare mein ek statement hai.

Figure — Classification — elliptic, parabolic, hyperbolic (discriminant test)

Topic ko yeh kyun chahiye: poori classification is baare mein hai ki $u$ ki shape kaise bend hone di jaati hai. Toh pehle humein woh object chahiye jis ki shape hum discuss karte hain.

1. Derivative — ek direction mein steepness

Agar $u$ sirf ek variable ka function hota, toh derivative $\frac{d u}{d x}$ answer karta: "height kitni fast change hoti hai jab main right step karta hoon?" Yeh curve ki slope hai.

Hum yahan derivative use karte hain kyunki ek PDE steepness aur curvature ke baare mein ek sentence hai — aur tum woh sentence nahi padh sakte jab tak tum nahi jaante "rate of change" ka matlab kya hai. Dekho Method of Characteristics jahan yeh slopes curves ban jaate hain.

2. Partial derivatives — jab do directions hon tab steepness

Ab $u$ dono $x$ aur $y$ par depend karta hai. Landscape par khade hokar, tum east step kar sakte ho ( $x$ badhao) ya north step kar sakte ho ( $y$ badhao), aur zameen dono taraf alag-alag rise karti hai. Ek partial derivative ek direction ko freeze karta hai aur doosre mein steepness measure karta hai.

Figure s02 — landscape par do directions. Blue contour lines $u$ ki level curves dikhate hain (equal height). Yellow point se, ek red arrow east point karta hai — uski slope $u_{x}$ hai; ek green arrow north point karta hai — uski slope $u_{y}$ hai. Picture clearly dikhati hai ki yeh same point par do alag steepnesses hain.

Topic ko yeh kyun chahiye: parent PDE $A u_{xx} + B u_{x y} + \dots$ inhi subscripts se bani hai. Agar tum $u_{x}$ nahi padh sakte toh equation nahi padh sakte.

3. Second partial derivatives — curvature

Ek slope lo, phir poocho "slope khud kitni fast change ho rahi hai?" Yeh ek second derivative hai. Ek landscape par yeh curvature measure karta hai — zameen upar bowl ki tarah bend hoti hai ya neeche dome ki tarah, aur kitni sharply.

Figure s03 — curvature ke teen flavours. Teen side-by-side slices: ek blue bowl ( $u_{xx} > 0$ , upar curve karta hai), ek red dome ( $u_{y y} < 0$ , neeche curve karta hai), aur ek yellow/green pair twist ( $u_{x y}$ ) dikhata hai — east-slope line north move karne par upar shift hoti hai, toh do directions coupled hain.

Definition Principal part (aur isse yeh naam kyun mila)

Ek second-order PDE ka principal part sirf uska highest-order piece hai: $A u_{xx} + B u_{x y} + C u_{y y}$ . "Principal" = "leading/sabse important" — yeh woh part hai jo type decide karta hai.

Deeper picture — yeh ek quadratic form hai. Agar tum har derivative ko direction ke liye ek symbol se replace karo — $u_{xx} \to α^{2}$ , $u_{x y} \to α β$ , $u_{y y} \to β^{2}$ (jahan $(α, β)$ plane mein ek direction hai) — principal part ban jaata hai $A α^{2} + B α β + C β^{2}$ . Yeh principal symbol hai: equation se attached ek purely geometric object, bilkul conic $A α^{2} + B α β + C β^{2} = const$ ki tarah shaped. Woh conic ellipse hai, parabola hai, ya hyperbola hai — yeh usi $B^{2} - 4 A C$ se decide hota hai — yahi reason hai ki teen PDE types woh names borrow karti hain. Sab kuch lower order ( $u_{x}, u_{y}, u$ , constants) is form ke bahar hota hai aur classification ignore karta hai.

4. Coefficients $A, B, C, D, E, F, G$ — equation ke "dials"

Parent equation hai $A u_{xx} + B u_{x y} + C u_{y y} + D u_{x} + E u_{y} + F u + G = 0.$

Har capital letter ek number (ya $x, y$ ka function) hai jo ek derivative term ko multiply karta hai — ek dial jo kehta hai "is tarah ke bending ka kitna hissa count hota hai."

Topic ko yeh kyun chahiye: discriminant $B^{2} - 4 A C$ hai. Tumhe koi bhi PDE dekh kar correctly yeh teen numbers pull out karni chahiye — including yeh realize karna ki ek missing second-order term ka matlab uska dial zero hai (jo exactly parabolic case ko drive karta hai, $Δ = 0$ ).

5. Quadratic formula — jahan $B^{2} - 4 A C$ paida hota hai

Ek quadratic $A m^{2} + B m + C = 0$ ki shape ki ek equation hai, jahan $m$ unknown hai. Jab $A \neq = 0$ ho, iske solutions quadratic formula se aate hain: $m = \frac{- B \pm B ^{2} - 4 A C}{2 A} .$

Square root ke neeche wala part, $B^{2} - 4 A C$ , discriminant kehlata hai.

Figure s04 — discriminant ka sign vs real roots ki sankhya. Teen parabolas $A m^{2} + B m + C$ : red wala horizontal axis ko do baar cross karta hai ( $Δ > 0$ , hyperbolic), yellow wala use sirf touch karta hai ( $Δ = 0$ , parabolic), blue wala upar float karta hai aur kabhi touch nahi karta ( $Δ < 0$ , elliptic). Jahan ek parabola axis se milti hai woh ek real root $m = d y / d x$ hai — ek real characteristic slope.

Topic ko yeh kyun chahiye: parent derivation mein characteristic slope $m = \frac{d y}{d x}$ satisfy karta hai exactly $A m^{2} + B m + C = 0$ . Toh "kitne real characteristic directions exist karte hain" wahi hai "is quadratic ke kitne real roots hain" wahi hai " $B^{2} - 4 A C$ ka sign kya hai." Yoh chain hi poori classification hai.

6. Characteristics — special curves (ek pehli jhalak)

Ek characteristic $x y$ -plane mein ek aisi curve hai jiske along PDE second derivative par apni pakad khoti hai — woh direction jiske along information (ek kink, ek wave front) travel karne di jaati hai. Inhi curves ki slopes Section 5 ke roots $m$ hain.

Poori machinery Method of Characteristics mein hai; yahan tumhe sirf itna jaanna chahiye ki yeh curves exist karti hain aur unki count real roots ki sankhya ke barabar hai.

7. Linear vs quasilinear — hum kis tarah ki equation classify kar sakte hain

Intuition Classification ke liye yeh distinction kyun matter karti hai — ek example ke saath

Discriminant test $A, B, C$ read karta hai aur ek type report karta hai. Lekin woh numbers kahan se aate hain?

Ek linear equation jaise Laplace ( $A = 1, B = 0, C = 1$ everywhere) ke liye, numbers sirf $x, y$ se fix hote hain, toh type solve karne se pehle tay ho jaati hai. Pehle classify karo, phir solve karo.
Ek quasilinear equation ke liye top dials mein $u$ hota hai — for instance $u u_{xx} + u_{y y} = 0$ mein $A = u$ hai, toh $Δ = - 4 u$ . Ab type solution ki value par depend karti hai: jahan bhi solution positive hai woh elliptic hai, jahan negative hai woh hyperbolic hai. Tum type nahi jaante jab tak (kuch had tak) answer nahi jaante.

Yahi poora reason hai ki distinction ko apna naam milta hai: linear PDEs ke liye classification equation ki property hai; quasilinear ones ke liye yeh solution ki property ho sakti hai. Yeh §4 ki position-dependence ko mirror karta hai lekin ek level aur deep le jaata hai. Linear vs Quasilinear PDEs mein aur zyada.

Prerequisite map

Neeche wala diagram Mermaid mein likha hai (flowcharts draw karne ka ek plain-text tarika). Har box ko ek concept ki tarah padho, aur har arrow ko "feeds into" ki tarah. Obsidian mein yeh ek real diagram ki tarah render hota hai; agar tumhe sirf text dikh raha hai, use top-to-bottom padho: baayein taraf ke foundations sab ek single number $B^{2} - 4 A C$ mein funnel karte hain, jo phir type decide karta hai.

Legend: F = landscape; P, S = uski slopes aur curvature; C = teen top dials; Q = quadratic step (sirf tab valid jab $A \neq = 0$ ); D = discriminant; CH = characteristic slopes; T = final verdict. Parent par wapas jao: Classification — discriminant test. Jis type par tum pohonchte ho woh dictate karta hai ki tumhe kaun sa data chahiye — dekho Well-posedness and Boundary Conditions.

Equipment checklist

Daayein taraf cover karo aur self-test karo:

Acronym PDE ka matlab kya hai?

Partial Differential Equation — kai variables ke ek unknown function aur uske partial derivatives ke liye ek equation.

u (x, y)

geometrically kya represent karta hai?

x y

-plane ke upar ek height — ek aisa landscape jiska shape PDE govern karta hai.

\partial

symbol ka matlab kya hai, aur plain

d

kyun nahi?

"Partial" derivative — ek variable ke respect mein differentiate karo jabki doosron ko fixed rakho; special

\partial

tumhe yaad dilata hai ki doosre variables frozen hain.

u_{x}

mein subscript tumhe kya batata hai?

Woh direction jisme tum step kiye (

x

= east) jabki doosra variable fixed rakha.

u_{x}

aur

u_{xx}

mein kya farq hai?

u_{x}

slope hai (first order, ek subscript);

u_{xx}

curvature hai (second order, do subscripts).

Sirf ek cross term

u_{x y}

kyun hai aur

u_{y x}

alag se kyun nahi?

Kyunki mixed partials commute karte hain (

u_{x y} = u_{y x}

) jab second partials continuous hon — Clairaut/Schwarz theorem.

Ek second-order PDE ka principal part kya hai, aur woh kaun sa geometric object form karta hai?

Uska highest-order piece

A u_{xx} + B u_{x y} + C u_{y y}

; derivatives ko directions se replace karne par quadratic form / principal symbol

A α^{2} + B α β + C β^{2}

milta hai — ek conic.

Discriminant mein kaun se coefficients enter karte hain, aur kaun ignore hote hain?

Sirf

A, B, C

(

u_{xx}, u_{x y}, u_{y y}

par dials);

D, E, F, G

ignore hote hain.

Kya

A, B, C

position ke functions ho sakte hain, aur iska kya result hota hai?

Haan — tab

Δ

(x, y)

ke saath vary karta hai aur type region se region mein change ho sakta hai (jaise Tricomi).

Test apply hone ke liye principal part ke baare mein kya true hona chahiye?

Yeh nondegenerate honi chahiye —

A, B, C

mein se kam se kam ek nonzero (ek genuine second-order PDE).

Discriminant likho.

Δ = B^{2} - 4 A C

B^{2} - 4 A C

ka sign kyun matter karta hai?

Yeh decide karta hai ki quadratic

A m^{2} + B m + C = 0

ke kitne real roots hain: positive → 2, zero → 1, negative → 0.

m

ke liye quadratic formula ko kaun si condition chahiye, aur agar

A = 0

ho toh kya karte ho?

A \neq = 0

(yeh

2 A

se divide karta hai); agar

A = 0

hai toh relabel karo ya type

Δ = B^{2} \geq 0

se seedha padho. Agar

A = C = 0, B \neq = 0

bhi ho toh type hyperbolic hai with characteristics coordinate lines ke along.

Quadratic ke real roots real-world mein kya ban jaate hain?

Characteristic curves ki slopes

d y / d x

Hyperbolic, parabolic, elliptic ke liye real characteristics ki sankhya?

2, 1, 0.

Linear vs quasilinear ek line mein, aur yeh kyun matter karta hai?

Linear: saare coefficients sirf

x, y

par depend karte hain (type solve karne se pehle fixed). Quasilinear: top-derivative coefficients

u

aur lower derivatives par depend kar sakte hain (type solution par depend kar sakti hai).