Liouville's theorem — phase space volume conservation: Foundations

2.1.14 · D1 · Physics › Analytical Mechanics › Liouville's theorem — phase space volume conservation

Yeh page — bilkul zero se — har woh symbol, word, aur picture build karta hai jis par parent note Liouville's Theorem depend karta hai. Isse upar se neeche padho: har idea agle ke liye ek brick hai.

1. Degrees of freedom, $N$

Simple words. Ek degree of freedom ek independent number hai jo tumhe dena padta hai taaki pata chale system kahan hai. Ek straight wire par sliding bead ko ek number chahiye (kitna aage). Ek flat table par bead ko do chahiye ( $x$ aur $y$ ). Hum is count ko $N$ kehte hain.

Picture. Ek control panel par kitne sliders ki zaroorat hogi machine ki position fully set karne ke liye — woh count $N$ hai.

Topic ko iske zaruarat kyun. Aage sab kuch — kitne coordinates, kitna bada space — $N$ se tay hota hai. Agar hum nahi jaante kitne knobs hain, toh kisi bhi cheez ke dimensions count nahi kar sakte.

2. Position coordinates $q_{i}$ aur momenta $p_{i}$

Simple words. Har degree of freedom $i$ ke liye (jahan $i = 1, 2, \dots, N$ run karta hai) hum do numbers rakhte hain:

$q_{i}$ — ek generalized position (kitna aage, kya angle, etc.),
$p_{i}$ — ek generalized momentum (roughly "kitni fast aur kitni heavy" us direction mein — plain particle ke liye mass times velocity).

Chhota subscript $i$ sirf ek label hai: $q_{1}$ pehli position hai, $q_{2}$ doosri, aur aise hi aage. Yeh koi power nahi hai; $q_{2}$ ka matlab hai "position number two", na ki " $q$ squared".

DONO position aur momentum kyun? Newton's laws second order hain — sirf yeh jaanna ki particle kahan hai uska future nahi batata, kyunki woh kisi bhi speed se move kar sakta hai. Tumhe pata hona chahiye woh kahan hai aur kaise move kar raha hai. Position + momentum sabse chhota package hai jo future ko unique banata hai.

3. Phase space — arena

Simple words. Un $2 N$ numbers ko lo aur unhe ek $2 N$ -dimensional space ke single point ke coordinates ki tarah treat karo. Woh space hai Phase Space. Ek point = poore system ka ek complete snapshot.

Picture. Ek particle ke liye 1D mein ( $N = 1$ ) phase space ek ordinary flat plane hai: horizontal axis position $q$ hai, vertical axis momentum $p$ hai. System is plane par ek dot "hai".

Figure — Liouville's theorem — phase space volume conservation

Topic ko iske zaroorat kyun. Liouville's theorem is space ke baare mein ek statement hai — specifically iske andar ke areas aur volumes ke baare mein. Phase space nahi, toh theorem state karne ko kuch nahi.

4. State as a vector $x$

Simple words. Har baar $2 N$ numbers likhna mushkil hai, isliye hum unhe ek bold symbol $x$ mein bundle karte hain: $x = (q_{1}, \dots, q_{N}, p_{1}, \dots, p_{N}) .$ Bold font ka matlab hai "yeh ek poori list (vector) hai, na ki single number". Ek picture: phase space ke origin se humari dot ki taraf pointing ek arrow.

Topic ko iske zaroorat kyun. Yeh flow laws ko compactly likhne deta hai: $2 N$ alag equations ki jagah hum kehte hain " $x$ aise move karta hai".

5. Dot: $\overset{q}{˙}$ , $\overset{p}{˙}$ , $\dot{x}$ — rate of change

Simple words. Kisi symbol ke upar ek dot ka matlab hai "yeh per unit time kitni tezi se change hota hai". Toh $\overset{q}{˙}$ position ki velocity hai, $\overset{p}{˙}$ woh rate hai jis par momentum change hota hai (jo ek force hai), aur $\dot{x}$ ek saath rates ka poora bundle hai.

Picture. Phase space ke har point par ek chhota arrow lagao jo kehta hai "agar dot yahan hai, toh yeh agli baar us taraf move karta hai". Yeh har jagah karo aur tum ek flow field paint karte ho — jaise arrows jo batate hain pani kahan kahan river mein kaunsi direction mein move karta hai. $\dot{x}$ wahi arrow hai.

Topic ko iske zaroorat kyun. Poora theorem is baare mein hai ki ek blob kaise flow karta hai. Dot-notation flow ki language hai.

6. Hamiltonian $H$

Simple words. $H$ ek single function hai — usually total energy — positions aur momenta ke terms mein likha gaya: $H (q, p)$ . Ise current $q$ 's aur $p$ 's do aur yeh ek number (energy) return karta hai.

Picture. $H$ ko phase space ke upar drawn ek landscape ki height ki tarah imagine karo: jahan energy high hai wahan hills, jahan low hai wahan valleys.

Topic ko iske zaroorat kyun. Yeh do equations flow ka law of motion hain. Liouville ke core mein jo magical cancellation hai woh directly inke mismatched signs se aata hai.

7. Partial derivative $\frac{\partial}{\partial x}$

Simple words. $H$ jaisi ek function kai variables par depend karti hai. Ek partial derivative $\partial H / \partial q_{i}$ poochhta hai: "agar main sirf $q_{i}$ ko thoda sa nudge karun aur baki sab freeze karun, toh $H$ kitni tezi se change hoti hai?" Yeh ek chosen direction mein energy landscape ka slope hai.

Picture. Energy-hill par khado, purely $q_{i}$ axis ke saath face karo, aur seedha aage ki ground ki steepness measure karo — sideways slopes ko ignore karte hue.

YEH tool kyun aur ordinary derivative $d / d x$ kyun nahi? Kyunki $H$ ke ek saath kai inputs hain. Ek ordinary derivative $d / d x$ sirf ek-input function ke liye sense deta hai. Curly $\partial$ exactly woh tool hai jo kehta hai "ek vary karo, baki hold karo" — precisely wahi jo Hamilton's equations demand karte hain.

8. Density $ρ$

Simple words. Hum exact state kabhi nahi jaante, isliye hum possible states ka ek cloud picture karte hain. $ρ (q, p, t)$ us cloud ki crowdedness hai: time $t$ par point $(q, p)$ ke paas kitne possible-states per unit phase-space volume hain.

Picture. Phase space par ek grey smudge — jahan states pile up hon wahan dark, jahan thin hon wahan faint. $ρ$ woh darkness hai.

Topic ko iske zaroorat kyun. Liouville ka sabse clean statement $ρ$ ke baare mein hai: ek moving dot ke saath ride karte hue, jo crowdedness tum dekhte ho woh kabhi change nahi hoti. Yeh $ρ$ Statistical Mechanics — Ensembles ka bridge hai.

9. Gradient $\nabla$ aur divergence $\nabla \cdot$

Simple words.

Gradient $\nabla ρ$ $ρ$ ke tamam partial-slopes ko ek arrow mein collect karta hai jo "upar ki taraf, denser ki taraf" point karta hai. Iska size batata hai density kitni steeply change hoti hai.
Divergence $\nabla \cdot \dot{x}$ measure karta hai ki flow field ek point par spread out ho raha hai (positive, ek source) ya squeeze in ho raha hai (negative, ek sink).

Divergence ke liye picture. Ek chhota box draw karo. Add up karo ki flow-arrow walls ke through kitna nikalta hai minus kitna enter karta hai. Agar zyada nikle, divergence positive hai (box ke contents blow apart ho rahe hain). Agar perfectly balance ho, divergence zero hai — ek incompressible flow.

YEH tool kyun? Ek flowing blob ki volume growth uske andar total divergence ke barabar hai. Toh "kya blob ka volume change hota hai?" ka jawab exactly "kya divergence zero hai?" se milta hai. Divergence woh mathematical naam hai us sawaal ka jo Liouville poochh raha hai.

10. Material (total) derivative $\frac{d}{d t}$

Simple words. Kisi quantity ko change hote dekhne ke do tarike hain:

$\frac{\partial ρ}{\partial t}$ — ek fixed point par still baitho aur cloud ko drift karte dekho.
$\frac{d ρ}{d t}$ — ek moving dot ke saath ride karo aur report karo ki woh kya dekhta hai.

Riding version dono clock-change aur yeh fact bundle karta hai ki tum naye territory mein move kar gaye: $\frac{d ρ}{d t} = \frac{\partial ρ}{\partial t} + \dot{x} \cdot \nabla ρ .$

Picture. Ek bridge par nailed thermometer (partial) versus ek thermometer jo river mein boat par float kar raha hai (total). Dono alag readings dete hain.

Topic ko iske zaroorat kyun. Liouville ka punchline, $d ρ / d t = 0$ , riding frame mein ek statement hai. Donon derivatives ke beech fark jaanna woh hai jo theorem ko quotable banata hai.

11. Poisson bracket ${,}$ (ek preview)

Simple words. Poisson bracket ${f, H}$ ek compact machine hai jo "jaise system $H$ ke under flow karta hai, $f$ kaise change hota hai" ko package karta hai. Iske poore workings ek sibling note mein hain; yahan sirf symbol se milo taaki parent ki line $\frac{\partial ρ}{\partial t} = - {ρ, H}$ stranger na lage. Yeh wahi Liouville statement hai flows ki natural language mein likhi hui.

Prerequisite map

Related roads is hub se bahar: Canonical Transformations (kyun volume-preservation structure mein built-in hai), Symplectic Geometry (woh deeper reason kyun mixed partials cancel hote hain), Poincaré Recurrence Theorem (finite conserved volume ka ek consequence).

Equipment checklist

Khud test karo — right side cover karo aur reveal karne se pehle jawab do.

Ek degree of freedom

N

kya hai?

Un independent numbers ki sankhya jo system ki position fix karne ke liye chahiye.

Hum

q_{i}

ke saath

p_{i}

bhi kyun track karte hain?

Newton's law second order hai — position akela future fix nahi karta; tumhe yeh bhi chahiye ki woh kaise move kar raha hai.

N

degrees of freedom ke liye phase space ki kitni dimensions hoti hain?

2 N

— ek axis per position aur ek per momentum.

\dot{x}

mein dot ka kya matlab hai?

Per unit time rate of change; state ka poora flow field.

Hamiltonian

H

usually kya hota hai?

Total energy positions aur momenta ka function likhke.

\frac{\partial H}{\partial p _{i}}

kya poochhta hai?

H

kitni tezi se change hota hai agar sirf

p_{i}

ko nudge kiya jaye aur baki sab held fixed rakhein.

Proof ke liye

H

ke mixed partials equal kyun hone chahiye?

Unki equality se do terms cancel ho jaate hain, jo flow ki divergence ko vanish hone par majboor karta hai.

ρ

kya represent karta hai?

Phase space mein possible states ke cloud ki crowdedness (density).

Divergence

\nabla \cdot \dot{x}

kya measure karta hai?

Kya flow spread out hota hai (positive) ya squeeze in hota hai (negative) ek point par; zero ka matlab incompressible.

\partial ρ / \partial t

aur

d ρ / d t

mein fark?

Partial = ek fixed point se dekhna; total = ek moving dot ke saath ride karna.

Liouville's theorem ek line mein?

Ek phase point par ride karte hue,

ρ

constant hai (

d ρ / d t = 0

); equivalently phase-space volume conserved hai.