Visual walkthrough — Gauss's law — integral form, choosing Gaussian surfaces

1.8.6 · D2 · Physics › Electromagnetism › Gauss's law — integral form, choosing Gaussian surfaces

---

Step 0 — Woh do pictures jinpar hume line one se pehle trust karna hai

Koi bhi symbol aane se pehle, do mental images par agree kar lete hain.

Ab har symbol jo hum use karenge woh in do pictures ke against earn kiya jaayega.

---

Step 1 — Ek patch, aur $\vec{E}\cdot d\vec{A}$ ka matlab asal mein kya hai

KYA HAI. Ek surface ka ek single tiny flat patch lo. Usmein do arrows hote hain:

• $\vec{E}$ — electric field arrow (direction jis taraf field push karti hai, length = strength).
• $d\vec{A}$ — patch ka area-vector: ek arrow jo patch ke perpendicular ($\perp$) khada hai, closed surface se bahar point karta hai, aur jiski length patch ki area ke barabar hai. Hum uski length $dA$ likhte hain, toh $|d\vec{A}|=dA$. Jab bhi aap plain scalar $dA$ dekhte ho uska matlab isi patch ki area hai — same object, bas uski length.

YEH DO KYUN. Flux hai "field jo surface face kar rahi hai." Facing measure karne ke liye, hume field arrow aur ek aise arrow ki zaroorat hai jo bataye ki patch kis taraf dekh raha hai. Woh doosra arrow hai $d\vec{A}$.

PICTURE. Figure mein, $\vec{E}$ angle $\theta$ par normal $d\vec{A}$ ki taraf aata hai. Sirf $\vec{E}$ ka woh shadow jo $d\vec{A}$ ki direction mein hai count hota hai — sideways wala hissa surface ke saath slide karta hai aur kuch nahi pakdta.

Dot product (upar define kiya hua) middle step automatically kar deta hai: $\vec{E}\cdot d\vec{A}=|\vec{E}||d\vec{A}|\cos \theta =|\vec{E}|dA\cos \theta$. $\cos \theta$ kyun aur $\sin \theta$ kyun nahi? Kyunki $\cos \theta$ tab $1$ hota hai jab field patch ko seedha head-on hit kare ($\theta =0$, full flux) aur $0$ jab field graze kare ($\theta =90^{\circ }$, zero flux). Yahi exactly rain-loop wala behaviour hai — isi liye dot product sahi tool hai, plain multiplication nahi.

---

Step 2 — Coulomb's law: woh ek field jis par hum shuru mein trust karte hain

KYA HAI. Space mein akela ek point charge $q$ rakh do. Coulomb's law hume uski field batata hai.

YAHAN SE KYUN SHURU KAREIN. Hum Gauss's law ko kuch nahi se derive nahi kar sakte — hum ise uss ek field law se bootstrap karte hain jo pehle se established hai: Coulomb.

PICTURE. Field arrows seedhe bahar radiate karte hain (agar $q>0$) jaise wheel ki spokes, distance ke saath patli hoti jaati hain.

• $r$ = charge se distance.
• $\hat{r}$ = length-one arrow jo $q$ se seedha door point karta hai.
• $\frac{1}{r^2}$ = field distance ke square ke saath fade hota hai — yeh sochte rehna, abhi yeh ek aisi area se takraane wala hai jo distance ke square ke saath barhti hai.
• ${\epsilon }_0=8.85\times 10^{-12}{\text{C}}^2{\text{N}}^{-1}{\text{m}}^{-2}$ = empty space kitna field ko "permit" karta hai.

---

Step 3 — Woh magical cancellation: sphere ke through flux

KYA HAI. $q$ ko radius $r$ ki ek sphere mein wrap karo, $q$ par centered. Poori sphere par flux add karo — yeh "ek closed surface par add karna" wahi hai jo integral sign par ring ka matlab hai:

SPHERE KYUN. Symmetry ke wajah se, sphere ka har point $q$ ke liye identical hai: same distance, aur wahan field seedha bahar point karti hai — sphere ke outward area-vector ke parallel. Toh $\theta =0$, $\cos \theta =1$ har jagah. Dot product $\vec{E}\cdot d\vec{A}$ phir $|\vec{E}|dA$ ke barabar ho jaata hai, aur $|\vec{E}|$ har patch par same hai, toh woh sum se bahar aa jaata hai.

PICTURE. Har red field arrow sphere ko head-on pierce karta hai; shell area $4\pi r^2$ hai.

Dot uss point tak survive karta hai jahan $\cos \theta =1$ hume use drop karne deta hai. Iss collision ko dekho: woh $r^2$ jisne field ko weak kiya wahi $r^2$ hai jisne sphere ko bada kiya. Woh bilkul perfectly divide ho jaate hain. Answer $q/{\epsilon }_0$ mein koi $r$ nahi — badi sphere weaker but wider field pakdti hai, aur total leakage kabhi nahi badlti. Yahi exact cancellation Gauss's law ka dhadakta hua dil hai.

---

Step 4 — Ek patch ko tilt aur stretch karo: solid angle prove karta hai ki sphere special nahi tha

KYA HAI. Sphere ko ek lumpy closed surface se replace karo. Usmein ek patch generally tilted hota hai (angle $\theta \ne 0$) aur sphere ke patch se zyada door hota hai.

KYUN. Hume dikhana hai ki koi bhi closed surface $q/{\epsilon }_0$ deta hai, sirf neat sphere nahi. Do effects ladte hain, aur hum dekhhte hain ki woh cancel ho jaate hain.

PICTURE. Area $dA$ ka ek tilted patch distance $r^{\prime }$ par, jahan $r^{\prime }$ = charge $q$ se is particular patch ki distance hai (yeh lumpy surface par patch-to-patch vary karti hai, sphere ke constant $r$ ke unlike). Uska head-on-facing part hai $dA\cos \theta$ (chhota — tilt se flux kho jaata hai). Lekin zyada door baithe hone se, field $1/r^{\prime 2}$ se weak hai... aur usse guzarne wala cone of lines ek solid angle $d\Omega$ define karta hai.

Distance $r^{\prime }$ aur tilt $\theta$ dono $d\Omega$ mein gaayab ho jaate hain. Lumpy closed surface same $q/{\epsilon }_0$ deta hai. Sphere sirf easy case tha; answer shape-blind hai.

---

Step 5 — Bahar wala charge: jo lines andar jaati hain woh bahar bhi nikalti hain

KYA HAI. Ab $q$ ko closed surface ke bahar poori tarah rakh do.

KYUN. Hume prove karna hai ki bahar ke charges zero net flux add karte hain — yahi woh cheez hai jo $Q_{\text{enc}}$ ko unhe ignore karne deti hai.

PICTURE. Bahar ke charge se aane wali ek field line ek taraf se closed surface mein dhus jaati hai (flux $-$, field wahan andar point karta hai, toh $\cos \theta <0$) aur doosri taraf se bahar nikal jaati hai (flux $+$). Andar, phir bahar: har line even number of times pierce karti hai, aur minus, plus ko cancel kar deta hai.

---

Step 6 — Ek saath kaafi saare charges: flux bas add ho jaata hai

KYA HAI. Real space mein kaafi saare charges hote hain. Unki fields superpose karti hain: $\vec{E}={\sum }_i{\vec{E}}_i$.

SUPERPOSITION KYUN KAAM KARTA HAI. Fields vectors ki tarah add hoti hain (Coulomb's law ka ek fact). Dot product aur sum ${\oint }_S$ dono linear hain — aap unhe $+$ signs par distribute kar sakte ho. Toh sum ka flux, fluxes ka sum hai.

PICTURE. Do charges andar, ek bahar. Andar wale har ek $q_i/{\epsilon }_0$ donate karte hain; bahar wala $0$ donate karta hai.

• $Q_{\text{enc}}$ = closed surface ke andar sealed charges ka grand total.
• Bahar ke charges silently $0$ ho jaate hain.
• Surface ki shape kabhi enter nahi ki. Ho gaya. $■$

---

Step 7 — Degenerate cases jinpar aapko trip nahi karna chahiye

KYA HAI. Woh corner cases jahan ek beginner khud ko second-guess karta hai.

KYUN. Ek law jo aap $q=0$ par, ya negative charge ke saath, ya surface par khud charge ke saath apply nahi kar sakte, woh law aap ne abhi poori tarah nahi seekha.

PICTURE. Teen chhote vignettes: empty closed surface, negative charge, charge jo surface par khud baitha hai.

• Empty surface ($Q_{\text{enc}}=0$). Net flux $=0$. Field lines andar se guzar sakti hain (bahar ke charges se) lekin jitni andar jaati hain utni bahar bhi aati hain. Zero net — zero field nahi.
• Andar negative charge ($q<0$). $\Phi =q/{\epsilon }_0<0$: lines andar point karti hain, net flux negative hai. Surface lines "peeta" hai.
• Charge bilkul surface par. Yeh well-posed setup nahi hai: charge ke through wala patch divergent field rakhta hai aur enclosed amount undefined hai. Rigorous fix hai ek limiting process — surface ko infinitesimal amount move karo taki charge cleanly andar ya cleanly bahar ho, phir limit lo. Practice mein: kabhi bhi Gaussian surface kisi point charge se guzar kar mat banao.
• Ek finite rod / off-centre charge ka field. Law phir bhi valid hai (hamesha sach!), lekin symmetry toot gayi hai, $E$ integral se bahar nahi aayega, aur aap $\vec{E}$ solve nahi kar sakte. Validity $\ne$ usability — wahi trap jaise Electric field of conductors problems mein hota hai.

---

Ek-picture summary

Upar sab kuch, compressed: charges ke saath ek lumpy closed surface jisme andar wale contribute karte hain aur bahar wale (net zero), field lines jo pierce karti hain unhe count kiya jaata hai, reference sphere par radius cancel hoti hai.

Recall Feynman retelling — ise apne plain words mein wapas bolo

Koi bhi closed surface imagine karo jo chahte ho. Field lines woh threads hain jo sirf plus charges par shuru hote hain aur minus charges par khatam hote hain — woh beech mein mid-air mein kabhi nahi rukते. Toh agar mujhe jaanna hai ki andar kitna charge trapped hai, main bas count karta hoon ki kitne threads net-bahar-poke karte hain uski skin ke through. Ek charge bahar ek taraf se ek line andar dhaansata hai aur doosri taraf se bahar, toh net zero hota hai — main use ignore kar sakta hoon. Ek charge andar ke saare threads escape kar jaate hain, toh woh hamesha same contribute karta hai, chahe main surface ko kaise bhi shape doon: maine yeh prove kiya sphere mein rakh kar, jahan field $1/r^2$ se weak hoti hai lekin sphere ki skin $r^2$ se barhti hai, aur dono cancel ho jaate hain toh radius gaayab ho jaati hai. Patches ko kisi bhi weird shape mein tilt aur stretch karo aur "solid angle" bookkeeping us cancellation ko intact rakhti hai. Bahut saare charges add karo — fields bas add hoti hain, toh fluxes bas add hote hain. Bahar-poke karne wale threads ka grand total trapped charge ko ${\epsilon }_0$ se divide karke barabar hota hai. Yahi poora law hai: net leakage caged charge count karta hai, aur kuch nahi.

Recall Quick self-test

Sphere ki radius kyun cancel hoti hai? ::: Field $\propto 1/r^2$ exactly utna hi shrink hoti hai jitna sphere ki area $\propto r^2$ barhti hai; product $q/{\epsilon }_0$ mein koi $r$ nahi. Kya bahar ka charge surface par $\vec{E}$ ko change karta hai? ::: Haan, pointwise — lekin uska net flux zero hai, toh woh $Q_{\text{enc}}$ se drop out ho jaata hai. Lumpy-surface case sphere case mein kyun reduce ho jaata hai? ::: Solid angle: tilt area ka $\cos \theta$ kho deta hai lekin distance $1/r^{\prime 2}$ compensate karta hai, aur $\oint d\Omega =4\pi$ hamesha hota hai. Zero charge enclose karne wali surface ke through net flux? ::: Zero — jo lines andar jaati hain woh bahar bhi aati hain. (Andar field zero hona zaruri nahi.) Gauss se finite rod ka $\vec{E}$ kyun nahi nikal sakte? ::: Koi symmetry nahi hai $E$ ko integral se bahar kheenchne ke liye; law valid hai lekin solvable nahi.