Visual walkthrough — Brachistochrone problem

4.10.14 · D2 · Maths › Advanced Topics (Elite Level) › Brachistochrone problem

---

Step 1 — Bead, wire, aur sawaal

KYA HAI. Do nailen, $A$ upar aur $B$ neeche aur ek taraf. Hum unke beech ek frictionless wire modte hain aur ek bead ko $A$ se rest se slide karne dete hain. Alag-alag wire shapes alag-alag travel times deti hain. Hum sabse tez shape chahte hain.

KYUN. Koi bhi maths karne se pehle, hume bilkul clear hona chahiye ki kya vary karta hai (wire ki shape) aur kya fixed hai (do endpoints, gravity). Hum ek poori curve choose kar rahe hain, koi single number nahi — yahi cheez ise mushkil banati hai.

PICTURE. Ek hi do nailon ke beech teen candidate wires: ek seedha ramp (magenta), ek halki sag (orange), aur ek steep early plunge (violet). Bead ko $A$ par ek chota dot draw kiya gaya hai.

Hum ise Calculus of Variations se jodte hain kyunki "sabse achha poora curve chunna" bilkul yahi iska kaam hai.

---

Step 2 — Time banta hai tiny slices se

KYA HAI. Wire ke ek microscopic piece ko zoom in karo. Us tiny piece mein bead ek tiny distance $ds$ move karta hai speed $v$ se, aur ek tiny time $dt$ leta hai.

KYUN. Hum "total time" directly nahi likh sakte — yeh poore path par depend karta hai. Lekin locally time aasaan hai: yeh hamesha distance ÷ speed hoti hai. Agar hum tiny piece ko sahi samajh lein, toh hum sab tiny pieces ko add up (integrate) kar ke total nikal sakte hain.

PICTURE. Curve par baitha ek blown-up right triangle: horizontal leg $dx$, vertical leg $dy$, aur hypotenuse $ds$ wire ke saath. Ek arrow bead ki speed $v$ us hypotenuse ke along label karta hai.

---

Step 3 — Pythagoras se slice length rewrite karna

KYA HAI. Hypotenuse $ds$ ko kuch aisa likhna jo purely $x$ mein ho.

KYUN. Jo functional hum chahte hain woh $x$ ke upar integral hona chahiye. Toh $ds$ (ek slanted length) ko $dx$ (ek horizontal step) aur curve ki steepness se express karna hoga.

PICTURE. Step 2 jaisa hi triangle, ab ratio $\frac{dy}{dx}$ slope ki tarah mark kiya gaya hai — per unit across kitne units neeche.

---

Step 4 — Speed force se nahi, energy se

KYA HAI. Depth $y$ par speed $v$ nikalna.

KYUN. Hum curved wire ke saath forces chase kar sakte the — ek nightmare. Iski jagah hum Conservation of Energy use karte hain: wire ki shape chahje kuch bhi ho, speed sirf is par depend karti hai ki bead kitna gira hai. Yeh woh single idea hai jo problem ko solvable banati hai.

PICTURE. Ek vertical "energy thermometer": top par saari energy potential hai (height); neeche, potential motion mein convert ho gayi hai (speed). Do beads depths $y_1<y_2$ par dikhate hain ki deeper wala faster ja raha hai.

---

Step 5 — Time functional assemble karna

KYA HAI. Steps 2–4 ko total time $T$ ke ek integral mein stack karna.

KYUN. Ab har tiny time $dt=ds/v$ ko $x$ aur curve ke slope mein likha ja sakta hai. Unhe sab add karne par ek single number milta hai jo poori curve $y(x)$ par depend karta hai — ek functional.

PICTURE. Curve kaafi saare thin vertical strips mein kata gaya; har strip apna $dt$ contribute karta hai, aur ek bracket dikhata hai unhe left-to-right mein $T$ mein sum hote.

Yeh $F$ ek Lagrangian ka role karta hai: woh quantity jiska integral hum optimise karte hain.

---

Step 6 — Beltrami shortcut (koi $x$ na hona ek gift kyun hai)

KYA HAI. Ek mushkil second-order equation solve karne ki jagah ek conserved quantity nikalna.

KYUN. Master tool Euler–Lagrange ek messy second-order ODE deta. Lekin jab $F$ mein koi explicit $x$ nahi hota, toh ek aisi quantity hoti hai jo optimal curve ke saath constant rehti hai — Beltrami identity. Yeh hume ek first-order equation free mein de deti hai, bilkul jaisa Step 4 mein energy conservation ne kiya tha.

PICTURE. Ek horizontal "conserved level" line: jab $\theta$/position curve ke saath change karta hai, combination $F-y^{\prime }{F}_{y^{\prime }}$ kabhi us line se nahi hatata, jabki $y$ aur $y^{\prime }$ dono wiggle karte hain.

$F=\frac{\sqrt{1+y^{\prime 2}}}{\sqrt{2gy}}$ plug in karke aur simplify karke (numerator $1$ ho jaata hai): $\frac{1}{\sqrt{2gy}\sqrt{1+y^{\prime 2}}}=C⟹y(1+y^{\prime 2})=k,k=\frac{1}{2g{C}^2}>0.$

---

Step 7 — ODE ko ek rolling wheel mein badalna

KYA HAI. $y(1+y^{\prime 2})=k$ solve karna aur shape pehchaanna.

KYUN — pehle variables separate karo. $y(1+y^{\prime 2})=k$ se slope isolate karte hain: $1+y^{\prime 2}=\frac{k}{y}\Rightarrow y^{\prime 2}=\frac{k-y}{y}\Rightarrow y^{\prime }=\frac{dy}{dx}=\sqrt{\frac{k-y}{y}}.$ Ab $dx$ akela ek side par laane ke liye flip karo (separation of variables — $x$ wali saari cheez left mein, $y$ wali saari cheez right mein, taaki dono sides apne aap integrate ho sakein): $dx=\sqrt{\frac{y}{k-y}}dy.$

• $\frac{k-y}{y}$ ::: isliye aata hai kyunki $y(1+y^{\prime 2})=k$ rearrange hoke yahi banta hai; yeh squared slope hai.
• $y/(k-y)$ ka square root ::: yahi integrate karna hai, lekin jaisa hai waisa awkward hai.

KYUN substitution. Blocker hai $\sqrt{y/(k-y)}$. Trick: $y$ aisa chunna ki $y$ aur $k-y$ dono sines/cosines ke perfect squares ban jaayein, kyunki phir messy square root ek clean tangent mein badal jaata hai. Half-angle identity $\frac{1-\cos \theta }{1+\cos \theta }={\tan }^2(\theta /2)$ bilkul yahi pattern hai — toh hum set karte hain $y=\frac{k}{2}(1-\cos \theta ),k-y=\frac{k}{2}(1+\cos \theta ),dy=\frac{k}{2}\sin \theta d\theta .$

PICTURE. Radius $r=k/2$ ka ek circle top line ke saath roll kar raha hai; uske rim par ek marked point curve trace karta hai jab angle $\theta$ ghoomta hai. Traced arc brachistochrone hai.

Substitute karke aur $\sin \theta =2\sin \frac{\theta }{2}\cos \frac{\theta }{2}$ use karke:

dx = \tan\!\tfrac\theta2 \cdot \frac{k}{2}\sin\theta\,d\theta = \frac{k}{2}(1-\cos\theta)\,d\theta.$$ $dx$ ko term by term integrate karte hain ($\int d\theta = \theta$, $\int\cos\theta\,d\theta = \sin\theta$): $$x = \frac{k}{2}(\theta - \sin\theta).$$ > [!formula] Cycloid > $r=k/2$ ke saath, > $$x = r(\theta - \sin\theta), \qquad y = r(1-\cos\theta).$$ > - $\theta$ ::: imaginary wheel kitna roll kar chuka hai (rolling angle). > - $r$ ::: wheel ka radius, is tarah set kiya gaya ki curve $B$ se guzre. > - $r\theta$ ::: wheel centre kitna move kiya hai; $-r\sin\theta$ / $-r\cos\theta$ pieces rim point ka us centre ke around wobbling hain. Yeh exactly ek [[Cycloid]] hai — light ke liye same "fastest path" idea ke liye [[Fermat's Principle]] dekho. --- ## Step 8 — Edge aur degenerate cases **KYA HAI.** Boundaries check karna taaki koi scenario reader ko surprise na kare. **KYUN.** Ek derivation tabhi trustworthy hai jab ends sahi behave karein. **PICTURE.** Teen mini-panels: (a) bilkul top par $y\to 0$ jahan curve vertical hai; (b) lowest point $\theta=\pi$; (c) comparison ke liye straight-line "cheat", clearly slower. ![[deepdives/dd-maths-4.10.14-d2-s08.png]] > [!example] Teeno limits > - **Top, $y\to0$:** $v=\sqrt{2gy}\to0$ (bead rest se shuru karta hai) aur slope $y'\to\infty$ — cycloid $A$ se *vertically* nikalti hai. Time finite rehta hai kyunki tezi se badhti speed steepness ko offset kar deti hai. > - **Bottom, $\theta=\pi$:** $y=2r$ (sabse gehra), slope $y'=0$ — curve momentarily flat hai, bead sabse fast. > - **Straight line (degenerate "wire"):** allowed hai lekin *optimal nahi*; yeh $v$ ko pehle chhota rakhta hai. Yeh sabse chhota $\int ds$ hai, kabhi nahi sabse chhota $\int ds/v$. --- ## Ek-picture summary ![[deepdives/dd-maths-4.10.14-d2-s09.png]] Ek figure mein poori story: rolling wheel (top), woh cycloid jo trace karta hai (sabse tez, violet), seedha chord (slower, magenta), tiny $ds$-triangle inset, aur labels $v=\sqrt{2gy}$, $y(1+y'^2)=k$, $x=r(\theta-\sin\theta),\,y=r(1-\cos\theta)$ wahan rakhe hain jahan har ek paida hua. > [!example] $T=\pi\sqrt{r/g}$ time kahan se aata hai > Cusp ($\theta=0$) se lowest point ($\theta=\pi$) tak ke descent ke liye, arc length aur speed parametrically likhte hain: > $$ds = r\sqrt{2(1-\cos\theta)}\,d\theta,\qquad v=\sqrt{2gy}=\sqrt{2gr(1-\cos\theta)}.$$ > Phir $(1-\cos\theta)$ factors $ds/v$ ke andar **cancel** ho jaate hain: > $$T=\int_0^\pi \frac{r\sqrt{2(1-\cos\theta)}}{\sqrt{2gr(1-\cos\theta)}}\,d\theta = \int_0^\pi \sqrt{\frac{r}{g}}\,d\theta = \pi\sqrt{\frac{r}{g}}.$$ > Woh cancellation hi tautochrone property hai: total sirf $r$ aur $g$ par depend karta hai. > [!recall]- Feynman ki retelling — poora walkthrough simple words mein > Ek bendy slide par ek marble, do fixed corners. Slide ko bahon mein kaato: har baal (length)÷(marble ki speed) leta hai. Pythagoras baal ki steepness se length deta hai; energy conservation speed deta hai sirf marble kitna gira hai se (heavy marbles matter nahi karte — mass cancel ho jaata hai). Sab baalon ko add karo → total time ek recipe $F$ ke roop mein. Woh recipe horizontal position directly kabhi mention nahi karti, aur yeh *laziness* ek gift hai: iska matlab hai ek khaas combination poore neeche jaane mein constant rehta hai (Beltrami), jo ek hard equation ko ek easy first-order mein badal deta hai, $y(1+y'^2)=k$. Ise solve karo aur — surprise — answer woh path hai jo ek rolling wheel par ek dot trace karta hai: ek cycloid. Yeh pehle almost seedha neeche plunge karta hai taaki pehle fast bane, phir across coast kare; farther-but-faster, shorter-but-slower se jeet jaata hai. Bonus: us curve par marble bottom tak *same* time leta hai chahe aap ise kahan bhi drop karo. > [!recall]- > $v=\sqrt{2gy}$ kahan se aata hai? ::: Energy conservation: $\tfrac12 mv^2=mgy$, mass cancel ho jaata hai. > Euler–Lagrange ki jagah Beltrami kyun use karte hain? ::: $F$ mein koi explicit $x$ nahi hai, toh $F-y'F_{y'}=C$ conserved hai aur ek first-order ODE deta hai. > Hume kaunsa ODE milta hai aur use kya solve karta hai? ::: $y(1+y'^2)=k$, cycloid $x=r(\theta-\sin\theta),\,y=r(1-\cos\theta)$ se solve hota hai. > Cusp se lowest point tak time? ::: $T=\pi\sqrt{r/g}$ ($ds/v$ mein $(1-\cos\theta)$ factors cancel ho jaate hain).