Visual walkthrough — Surface integrals — scalar and vector (flux)

4.4.31 · D2 · Maths › Multivariable Calculus › Surface integrals — scalar and vector (flux)

Hum is ek line tak pahunchenge: ${\iint }_S\mathbf{F}\cdot d\mathbf{S}={\iint }_D\mathbf{F}(\mathbf{r}(u,v))\cdot ({\mathbf{r}}_u\times {\mathbf{r}}_v)dudv.$ Abhi iske kisi bhi symbol ki chinta mat karo — including $d\mathbf{S}$, jise hum Step 7 mein carefully build karte hain. Hum unse ek-ek karke milenge.

---

Step 1 — Ek flat parameter rectangle hi woh cheez hai jo hum samajhte hain

KYA. Ek flat plane se shuru karo jiske do directions hum $u$ (horizontal) aur $v$ (vertical) kehte hain. Iske andar ek region $D$ chuno — ise graph paper ki ek sheet samjho. Yahi hamara home base hai.

KYUN. 3D mein floating ek curved surface par directly integrate karna hopeless hai — ek curved sheet par koi clean "left-to-right, bottom-to-top" order nahi hota. Lekin ek flat rectangle ko hum pehle se sweep karna jaante hain: woh exactly ek double integral hai. Toh poora plan yeh hai: saari mushkil kaam flat sheet par karo, phir correct karo ki surface ne use kitna distort kiya.

PICTURE. Neeche wala green square flat sheet ka ek tiny tile hai. Iska width $u$ ka ek tiny amount hai jise hum $du$ likhte hain, aur height $v$ ka ek tiny amount hai jise hum $dv$ likhte hain. "$d$" ka matlab sirf bahut chota tukda hai. Iska area honestly $dudv$ hai — ek plain rectangle.

---

Step 2 — Map $\mathbf{r}$ flat sheet ko curved surface par lift karta hai

KYA. Ek parametrization ek rule $\mathbf{r}(u,v)$ hai jo har flat point $(u,v)$ ko leta hai aur batata hai ki woh 3D space mein kahan land karta hai: $\mathbf{r}(u,v)=(\underset{\text{left–right}}{\underbrace{x(u,v)}},\underset{\text{forward–back}}{\underbrace{y(u,v)}},\underset{\text{up–down}}{\underbrace{z(u,v)}}).$ $x,y,z$ mein se har ek ek plain number hai jo do dials $u$ aur $v$ par depend karta hai.

KYUN. Yahi bridge hai. Do dials $u,v$ ghuma kar surface $S$ ke har point par ja sakte hain. Bold $\mathbf{r}$ ("position vector") sirf origin se us landed point tak ka ek arrow hai — teen coordinates ek saath bundle kiye hue.

PICTURE. Left par flat green tile $\mathbf{r}$ ke zariye right par curved surface ke ek warped patch par carry ho jaata hai. Ab woh flat rectangle nahi raha — woh bend aur stretch ho gaya hai.

---

Step 3 — Do tangent arrows dikhate hain ki tile kis taraf stretch hui

KYA. Surface par ek point par baithho. Sirf $u$ nudge karo ($v$ frozen rakho): landing point ek curve ke saath slide karta hai. Iska velocity — woh direction aur speed jis se woh $u$ ke per unit move karta hai — partial derivative hai: ${\mathbf{r}}_u=\frac{\partial \mathbf{r}}{\partial u}=(\frac{\partial x}{\partial u},\frac{\partial y}{\partial u},\frac{\partial z}{\partial u}).$ Isi tarah sirf $v$ nudge karne se ${\mathbf{r}}_v=\partial \mathbf{r}/\partial v$ milta hai.

KYUN yahan derivative? Hum jaanna chahte hain ki map ek infinitesimal tile ko kaise distort karta hai. "Position ka rate of change jab main ek dial ghuma raha hoon" exactly ek derivative ki job description hai. Partial word (curly $\partial$) ka matlab sirf "doosre ko freeze karte hue ek dial ke respect mein differentiate karo" hai. Hum derivative choose karte hain — say, average nahi — kyunki distortion ek local, instantaneous cheez hai.

PICTURE. ${\mathbf{r}}_u$ (magenta) $u$-curve ke along point karta hai; ${\mathbf{r}}_v$ (violet) $v$-curve ke along point karta hai. Woh do edges hain jis par little tile map hui — dono surface ke tangent hain (woh use graze karte hain, kabhi nahi chodते).

---

Step 4 — Tile ek tiny parallelogram ban jaata hai, aur humein iska area chahiye

KYA. Mapped tile (first order tak) parallelogram hai jiske edge vectors ${\mathbf{r}}_{u}du$ aur ${\mathbf{r}}_{v}dv$ hain. Yahan ${\mathbf{r}}_{u}du$ ka matlab hai "tangent arrow ${\mathbf{r}}_u$ us length tak shrink kiya jitna step $du$ actually produce karta hai."

KYUN. Itne tiny tile ke upar, curved patch aur flat parallelogram jo iske do edge arrows se span hota hai, indistinguishable hain. Curvature sirf second order par matter karta hai — tiny se bhi chota — toh hum use ignore karte hain. Hamara poora sawaal — "ek flat tile kitna real area ban jaata hai?" — ab yeh hai ki "is parallelogram ka area kya hai?"

PICTURE. Left par area $dudv$ ka flat tile; right par iska image, ek slanted parallelogram. Dono mein tiles ki ek jaisi sankhya hai lekin alag areas — wahi gap poori story hai.

---

Step 5 — Cross product us parallelogram ka area measure karta hai

KYA. Edge vectors $\mathbf{a}$ aur $\mathbf{b}$ wale parallelogram ka area unke cross product ki length hai: $\text{area}=\Vert \mathbf{a}\times \mathbf{b}\Vert =\Vert \mathbf{a}\Vert \Vert \mathbf{b}\Vert \sin \theta .$ Yahan $\theta$ edges ke beech ka angle hai, aur $\Vert \cdot \Vert$ ka matlab "ki length" hai.

KYUN cross product, sabhi tools mein se? Humein ek saath do cheezein chahiye: parallelogram का area aur, baad mein, woh direction jo iske seedha baahir point kare. Cross product dono ek hi object mein deliver karta hai — iska length area hai, iska direction dono edges ke perpendicular hai. Koi aur product dono nahi karta. $\sin \theta$ exactly sahi hai: yeh hai (ek edge) $\times$ (woh height jo doosri edge $\theta$ se tilt hone par pahunchti hai) = base $\times$ height.

Hamare edges $\mathbf{a}={\mathbf{r}}_{u}du$, $\mathbf{b}={\mathbf{r}}_{v}dv$ daalo. Kyunki $du,dv$ plain positive numbers hain woh bahar aa jaate hain: $\Vert {\mathbf{r}}_{u}du\times {\mathbf{r}}_{v}dv\Vert =\underset{\text{stretch factor}}{\underbrace{\Vert {\mathbf{r}}_u\times {\mathbf{r}}_v\Vert }}\underset{\text{flat tile area}}{\underbrace{dudv}}.$

PICTURE. Cross product ${\mathbf{r}}_u\times {\mathbf{r}}_v$ (orange) patch ke perpendicular khada hai; shaded parallelogram ka area iska length ke equal hai. Jab do edges almost parallel hoti hain ($\theta \to 0$), $\sin \theta \to 0$ aur area collapse ho jaata hai — right par thin sliver dekho.

---

Step 6 — Arrow ko ek unit normal mein badhalo aur poocho "kitna flow bahar jaata hai?"

KYA. Orange arrow ko uski apni length se divide karo taaki length 1 ka pure-direction arrow mile, jo unit normal hai: $\hat{\mathbf{n}}=\frac{{\mathbf{r}}_u\times {\mathbf{r}}_v}{\Vert {\mathbf{r}}_u\times {\mathbf{r}}_v\Vert }.$ Chota hat ($\hat{}$) hamesha matlab "length one" hota hai.

KYUN. Ab ek vector field $\mathbf{F}$ laao — space ke har point par yeh ek arrow hai, maan lo fluid ki velocity. Hamare tiny patch se, sirf woh part of $\mathbf{F}$ jo $\hat{\mathbf{n}}$ ke saath point karta hai actually surface cross karta hai. Sideways flow sirf patch ke saath slide karta hai aur kuch bhi through nahi jaata. "$\hat{\mathbf{n}}$ ke along part" extract karne ke liye hum dot product $\mathbf{F}\cdot \hat{\mathbf{n}}$ use karte hain — woh ek tool jo ek arrow ko doosre ke direction par project karta hai (yeh $\Vert \mathbf{F}\Vert \cos \alpha$ return karta hai, jahan $\alpha$ $\mathbf{F}$ aur $\hat{\mathbf{n}}$ ke beech ka angle hai).

PICTURE. $\mathbf{F}$ (magenta) patch par ek angle se hit karta hai. Iska $\hat{\mathbf{n}}$ (orange, length 1) par shadow through-part hai; baaki violet component sideways slide karta hai aur kuch bhi through nahi le jaata.

---

Step 7 — Square root cancel ho jaata hai: punchline

KYA. Pehle woh object name karo jise hum bana rahe hain. Vector area element $d\mathbf{S}$ patch ka area aur uski outward direction ek single arrow mein bundle karta hai:

Ab Step 6 ke do pieces multiply karo. Vector dot product (vector-dot-vector) ke liye $[\cdot ]$ likho aur ordinary scalar multiplication juxtaposition se likho: $(\mathbf{F}\cdot \hat{\mathbf{n}})dS=[\mathbf{F}\cdot \frac{{\mathbf{r}}_u\times {\mathbf{r}}_v}{\Vert {\mathbf{r}}_u\times {\mathbf{r}}_v\Vert }](\Vert {\mathbf{r}}_u\times {\mathbf{r}}_v\Vert dudv).$ Is line mein brackets andar wala $\cdot$ do arrows ka dot product hai (jo ek number deta hai); bracket phir multiply hota hai (plain number times number) factor $\Vert {\mathbf{r}}_u\times {\mathbf{r}}_v\Vert dudv$ se. Woh scalar factor $\hat{\mathbf{n}}$ ke bottom mein aur $dS$ ke top mein hai, toh exactly cancel ho jaata hai: $(\mathbf{F}\cdot \hat{\mathbf{n}})dS=\mathbf{F}\cdot ({\mathbf{r}}_u\times {\mathbf{r}}_v)dudv.$

KYUN yeh beautiful hai. Ek scalar surface integral ke liye tum messy root $\Vert {\mathbf{r}}_u\times {\mathbf{r}}_v\Vert =\sqrt{\cdots }$ compute karne mein stuck ho. Flux ke liye woh evaporate ho jaata hai — tumhe kabhi square root lena hi nahi padta. Bas $\mathbf{F}$ ko raw cross product ke saath dot karo. Toh vector area element simplify ho jaata hai: $d\mathbf{S}=\hat{\mathbf{n}}dS=({\mathbf{r}}_u\times {\mathbf{r}}_v)dudv,$ area aur direction koi root dikhaye bina carry karta hai.

PICTURE. Left: $\hat{\mathbf{n}}$ (length 1) times ek fat patch. Right: cancellation ke baad, ek honest arrow ${\mathbf{r}}_u\times {\mathbf{r}}_v$ jiska length hi area hai aur jiska direction hi orientation hai — koi denominator survive nahi karta.

---

Step 8 — Unit sphere par sanity check (saare pieces ek saath)

Setup — spherical parameters $\varphi ,\theta$. Ek sphere ke liye do dials $u,v$ ko do natural spherical angles mein rename kiya jaata hai:

• $\varphi$ (polar angle, "north pole se kitna neeche"), range $\varphi \in [0,\pi ]$: $\varphi =0$ north pole hai, $\varphi =\pi$ south pole.
• $\theta$ (azimuth, "equator ke around kitna"), range $\theta \in [0,2\pi ]$.

Toh yahan $(u,v)=(\varphi ,\theta )$, aur Steps 3–5 ke tangent arrows ${\mathbf{r}}_{\varphi }=\partial \mathbf{r}/\partial \varphi$ (ek meridian ke along southward point karta hai) aur ${\mathbf{r}}_{\theta }=\partial \mathbf{r}/\partial \theta$ (latitude ke ek circle ke along eastward point karta hai) ban jaate hain. Baaki sab exactly wahi machine hai jo humne abhi banaayi.

KYA. $\mathbf{F}=(x,y,z)$ aur outward unit sphere lo. Parametrize karo $\mathbf{r}(\varphi ,\theta )=(\sin \varphi \cos \theta ,\sin \varphi \sin \theta ,\cos \varphi )$. Compute karte hain ${\mathbf{r}}_{\varphi }\times {\mathbf{r}}_{\theta }=\sin \varphi \mathbf{r}$. Phir, sphere par $\mathbf{F}=\mathbf{r}$ aur $\Vert \mathbf{r}\Vert =1$ hone se, $\mathbf{F}\cdot ({\mathbf{r}}_{\varphi }\times {\mathbf{r}}_{\theta })=\mathbf{r}\cdot (\sin \varphi \mathbf{r})=\sin \varphi \Vert \mathbf{r}{\Vert }^2=\sin \varphi .$ ${\iint }_S\mathbf{F}\cdot d\mathbf{S}={\int }_0^{2\pi }{\int }_0^{\pi }\sin \varphi d\varphi d\theta =2\pi \cdot 2=4\pi .$

KYUN yeh example. Yahan $\mathbf{F}$ radially outward point karta hai, har jagah surface ke seedha through, toh har patch fully contribute karta hai — flux maximum aur positive hona chahiye. Woh $4\pi$ hai, exactly sphere ka surface area, jo sense banata hai: unit-speed outward flow through unit-area = area. Divergence Theorem se cross-check: $\nabla \cdot \mathbf{F}=3$ times ball volume $\frac{4}{3}\pi$ gives $4\pi$. ✔ ($\varphi$ phir $\theta$ order ${\mathbf{r}}_{\varphi }\times {\mathbf{r}}_{\theta }$ ko outward point karta hai — right-hand rule woh orientation confirm karta hai jo problem chahta hai.)

PICTURE. Radial arrows $\mathbf{F}$ sphere ko dead-on pierce karte hue, har normal $\hat{\mathbf{n}}$ flow ke saath aligned — kuch bhi sideways slide nahi karta.

---

Ek-picture summary

Ek nazar mein: ek flat tile $dudv$ → $\mathbf{r}$ se map hua → do tangent edges ${\mathbf{r}}_u,{\mathbf{r}}_v$ → cross product area aur direction deta hai → field ko dot se project karo → root cancel ho jaata hai → flat $D$ par integrate karo.

Recall Feynman retelling — poora walkthrough kisi dost ko explain karo

Socho tum ek bumpy fishing net ko ek waterfall ke neeche pakde ho aur poocha "per second kitna paani through jaata hai?"

Net curved aur reason karna awful hai, toh main ek flat graph paper ki sheet uske paas rakhta hoon aur pretend karta hoon ki paper par har chhota square net par ek bumpy patch correspond karta hai (map $\mathbf{r}$). Lekin ek flat square ek stretched, tilted patch par land karta hai, toh main iske do edges measure karta hoon — woh arrows jo dikhate hain ki patch kitna move hua jab maine ek graph-paper direction ek time par slide ki (tangent vectors ${\mathbf{r}}_u,{\mathbf{r}}_v$). Us tilted patch ka real area aur woh direction jis taraf woh ek hi shot mein face karta hai paane ke liye, main do edges cross karta hoon (cross product): iska length area hai, iska arrow net se seedha baahir point karta hai. Ab waterfall: sirf woh paani jo us outward arrow ke saath aim karta hai actually cross karta hai — sideways splash sirf off slide ho jaata hai — toh main paani ki velocity ko outward direction par shadow karta hoon (dot product). Jab main sab likh deta hoon, "length one tak shrink karo" division aur "area se multiply karo" dono same length use karte hain, aur woh cancel ho jaate hain — toh flux sirf velocity dotted with raw cross product hai, graph paper ke har square par sum kiya hua. Koi square roots nahi, aur agar maine net flip kiya toh answer sign flip kar leta hai.

Recall Quick checks

Tangents ke liye derivative kyun? ::: Yeh map ki instantaneous stretch measure karta hai jab ek parameter change hoti hai. Cross product kyun aur kuch nahi? ::: Yeh area (iska length) aur outward direction (iska arrow) simultaneously deta hai. Root flux ke liye cancel kyun hota hai lekin scalar integrals ke liye nahi? ::: $\hat{\mathbf{n}}$ length se divide karta hai; $dS$ usse multiply karta hai — woh sirf flux mein milte hain. Ek singular point par kya hota hai jahan ${\mathbf{r}}_u\parallel {\mathbf{r}}_v$? ::: Cross product zero hai; tile collapse ho jaati hai aur koi area contribute nahi karti. ${\mathbf{r}}_u\times {\mathbf{r}}_v$ vs ${\mathbf{r}}_v\times {\mathbf{r}}_u$ kaise choose karte hain? ::: Right-hand rule: ungliyan increasing $u$ ke along curl karke increasing $v$ ki taraf jaati hain, thumb normal hai; parameter order choose karo taaki thumb woh direction point kare jaise problem maangti hai.