Buckling — Euler column buckling load derivation: Visual walkthrough

3.6.6 · D2 · Physics › Spacecraft Structures & Systems Engineering › Buckling — Euler column buckling load derivation

Hum poore time ek hi sawaal ke peechhe hain:

Kitne pushing force par ek perfectly straight column suddenly sideways bow karna prefer karta hai?

Step 0 — Pencil-on-its-tip waali picture (yeh "buckling" exist hi kyun karta hai)

KYA. Ek patla ruler lo. Uske dono ends ko ek doosre ki taraf dhakelo, gently. Kuch nahi. Zyada dhakelo — kisi ek sudden threshold par woh ek taraf spring kar jaata hai ek smooth bow mein. Woh chota aur crush nahi hua; woh sideways gaya.

WHY yahaan se shuru karein. Kisi bhi maths se pehle, aapko yeh phenomenon believe karna hoga: threshold force se neeche straight shape stable hai (nudge karo, woh wापas aata hai); threshold se upar straight shape unstable hai (nudge karo, woh bhaag jaata hai aur bend ho jaata hai). Hum exactly usi threshold force ko dhundh rahe hain.

PICTURE. Neeche, left column stable hai (ek nudge khatam ho jaata hai, arrows straight ki taraf point karte hain). Right column threshold se aage hai — ek nudge barhta hai (arrows door point karte hain). Beech mein jo threshold hai, wahi number hum chahte hain.

Step 1 — Bent column draw karo aur har coordinate ka naam rakho

KYA. Column ko khada karo. $x$ ko $0$ se (bottom pin par) $L$ tak (top pin par) run karne do. Har height $x$ par, column sideways $y (x)$ jitna bow kar chuka hai. Ends pinned hain: jagah par pakde hue hain lekin swivel karne ke liye free hain (jaise door hinge), isliye $y (0) = 0$ aur $y (L) = 0$ .

KYU. Buckling hai hi ek nonzero $y (x)$ ka janam. Isliye hum us sideways shape ko ek naam aur coordinate frame dete hain — tabhi hum pooch sakte hain ki kaunsi force use allow karti hai.

PICTURE. $x$ vertical axis hai (blue), $y$ horizontal hai (yellow). Dono ends par pins $y = 0$ par pinned hain; beech mein column bow karta hai. Ek single smooth hump notice karo — abhi koi wiggles nahi, woh humne abhi earn nahi kiya.

$sideways offset y (height along column x)$

$x$ :: ek position, $0 \to L$ run karta hai.
$y (x)$ :: us height par sideways displacement. Yeh woh unknown shape hai jiske liye hum solve karte hain.

Step 2 — Bending moment kahaan se aata hai? (feedback loop)

KYA. Column ko kisi height $x$ par kaato aur upar waale piece ko dekho. Load $P$ abhi bhi original centre-line ke saath seedha neeche push karta hai. Lekin cut par material sideways $y (x)$ drift kar chuka hai. Ek sideways distance par act karne wali force cut ke baare mein ek twisting effect produce karti hai — ek bending moment.

KYU. Yahi buckling ka poora raaz hai. Push $P$ column ko tabhi bend karta hai jab column pehle se bend ho chuka ho — offset $y$ lever arm hai. Zyada bend ⇒ zyada moment ⇒ aur zyada bend. Yeh positive feedback hai, aur positive feedback hi straight state ko unstable banata hai.

PICTURE. Red arrow $P$ hai jo dashed original axis ke neeche act kar raha hai. Green segment lever arm $y (x)$ hai — force jahan act karti hai aur material ab jahan hai, unke beech horizontal gap. Curved red arrow woh moment hai jo yeh produce karta hai, bow ko badhane wali direction mein curl karta hua.

$M (x) = sign: moment bend badhata hai - axial push (N) P lever arm (m) y (x)$

Minus sign bookkeeping hai: hamare sign convention ke saath ek positive deflection ek aisa moment produce karta hai jo mathematical sense mein beam ko axis ki taraf wापas curve karta hai — lekin physically effect woh runaway feedback hai jo upar hai. (Iska kaam Step 4 ki equation ko ek wave equation banane ka hai, runaway exponential nahi — hum dekhenge kyun yeh matter karta hai.)

Step 3 — Ek beam moment ka jawab kaise deta hai (material ka reply)

KYA. Ek beam bending ka resist karta hai. Euler-Bernoulli Beam Theory se, ek beam kitna curve hota hai woh apply hone wale moment ke proportional hota hai: $E I \frac{d ^{2} y}{d x ^{2}} = M (x)$

KYU exactly yeh tool — aur kyun second derivative? Hume ek aisa number chahiye jo measure kare "shape kitni sharply bend ho rahi hai" har height par. First derivative $d y / d x$ sirf slope hai (column kitna tilted hai) — ek straight tilted line mein slope hota hai lekin bend nahi. Jo cheez bend measure karti hai woh yeh hai ki slope khud kitni tezi se change hota hai: second derivative $d^{2} y / d x^{2}$ , yaani curvature. Isliye exactly ek second derivative appear hota hai aur first nahi: curvature, tilt nahi, woh hai jiske against moment fight karta hai.

PICTURE. Teen sample shapes: ek straight tilt (slope, lekin $d^{2} y / d x^{2} = 0$ , koi curvature nahi), ek gentle bow (small curvature), ek sharp bow (large curvature). Bow jitna sharp, $d^{2} y / d x^{2}$ utna bada, aur equation se moment utna bada chahiye.

$Young’s modulus: material stiffness E second moment of area: material kitna door centre se baitha hai I curvature \frac{d ^{2} y}{d x ^{2}} = applied moment M$

$E$ :: material se. Bada $E$ , stiffer cheez, curve karna mushkil.
$I$ :: Second Moment of Area se. Material ko centre se bahar dhakelo aur $I$ shoot up karta hai — tube ek rod ko beat karta hai.
Saath mein $E I$ beam ki bending stiffness hai. Note: koi bhi yield strength nahi — buckling ek stiffness ki kahani hai, strength ki nahi.

Step 4 — Dono facts ko joodo: governing equation

KYA. Step 2 ne kaha $M = - P y$ . Step 3 ne kaha $E I y^{''} = M$ . Dono ko equal set karo: $E I \frac{d ^{2} y}{d x ^{2}} = - P y ⟹ \frac{d ^{2} y}{d x ^{2}} + k^{2} y = 0, k^{2} \equiv \frac{P}{E I}$

KYU. Humne moment ka cause (offset par act karne wala load) ko moment ke against beam ka response ke saath combine kiya. Jo bhi shape $y (x)$ survive karti hai, use yeh single equation follow karni hogi. Humne poore clump $P / E I$ ko $k^{2}$ rename kiya sirf algebra clean rakhne ke liye — $k$ ki units 1/length hain, isliye $k x$ ek pure number hai (ek angle), jo hume aage chahiye hoga.

PICTURE. Yeh equation kehti hai: har point par curvature wahaan ki deflection ke negative ke proportional hai. Jahan column far right bow karta hai (positive $y$ ), woh left ki taraf curve karta hai; jahan yeh centre cross karta hai, woh momentarily straight hai. Woh "hamesha middle ki taraf wापas curve karna" ek wave ki fingerprint hai. Ek spring par mass se compare karo, jahan acceleration hamesha ghar ki taraf point karta hai — same equation, lekin yahaan independent variable height $x$ hai, time nahi.

$\frac{d ^{2} y}{d x ^{2}} + = P / E I (tuning knob) k^{2} y = 0$

$k^{2} = P / E I$ :: kitna hard push kiya ( $P$ ) ko beam kitni stiff hai ( $E I$ ) ke against bundle karta hai. $P$ badhao toh $k$ badhata hai — wave tighter hoti hai. Iske upar nazar rakho; yeh sab decide karta hai.

Step 5 — Is equation ki allow karda shapes

KYA. $y^{''} + k^{2} y = 0$ ka har solution "wavenumber" $k$ ki sine aur cosine wave ka mix hai: $y (x) = A sin (k x) + B cos (k x)$

KYU yeh functions? Hume aise shapes chahiye the jinki curvature minus themselves ho (Step 4). Sine aur cosine akele functions hain jisme yeh property hai: $sin$ ko do baar differentiate karo aur aapko $- sin$ milta hai. Isliye buckled shape wavy hone par majboor hai — choice se nahi, balki equation ki wajah se. $A$ aur $B$ unknown dial settings hain (amplitudes) jinhe hum pins se pin down karenge.

PICTURE. Ek sine zero se start hota hai; cosine apne peak se start hota hai. General solution dono ka koi blend hai. Unki combined wavelength $k$ se set hoti hai: bada $k$ (hard push) wave ko shorter squeeze karta hai.

$y (x) = sine amount A sin (k x) + cosine amount B cos (k x)$

Step 6 — Pins zyaadatar shapes ko veto karte hain (boundary conditions)

KYA. Pins force karte hain $y = 0$ dono ends par.

Bottom par, $x = 0$ : $y (0) = A sin 0 + B cos 0 = B = 0.$ Isliye $B = 0$ — cosine dead hai.
Top par, $x = L$ : $y (L) = A sin (k L) = 0.$

KYU. Ek pin physically end ko centre-line par pin karta hai: woh sideways move nahi kar sakta. Cosine $x = 0$ par $1$ hota hai, isliye woh pin pe ek bulge daalega — impossible, isliye $B = 0$ . Jo bacha woh ek pure sine hai, $y = A sin (k x)$ . Top pin par sine ko bhi zero par wापas aana hoga.

Ab $x = L$ par raaste ka fork:

Agar $A = 0$ bhi hai, toh $y (x) = 0$ har jagah — straight column. Hamesha valid solution, lekin boring.
Ek genuinely bent column ke liye hume $A \neq = 0$ chahiye, jo force karta hai: $sin (k L) = 0$

PICTURE. Sirf woh sine waves allowed hain jo exactly $x = L$ par zero hit karti hain. Ek wave jiska "natural" zero $L$ se pehle ya baad land karta hai, woh vetoed hai — woh top pin satisfy nahi kar sakti. Allowed waves quantised hain.

$B = 0$ :: cosine ko khatam karta hai (pin par bulge nahi ho sakta).
$sin (k L) = 0$ :: bent shape ke liye survival condition. Yahaan se special loads aate hain.

Step 7 — Sirf special loads survive karte hain (eigenvalue staircase)

KYA. $sin (angle) = 0$ tabhi hota hai jab angle $π$ ka pura multiple ho (wahaan sine zero cross karta hai: $0, π, 2 π, \dots$ ). Isliye: $k L = nπ, n = 1, 2, 3, \dots ⟹ k = \frac{nπ}{L} .$ Rename $k^{2} = P / E I$ ko undo karo: $P = \frac{n ^{2} π ^{2} E I}{L ^{2}}$

KYU. Pins ek bent shape sirf discrete loads par allow karte hain — ek har whole number of half-waves ke liye jo ends ke beech fit hoti hain. Mathematicians ise exactly eigenvalue problem kehte hain: special values ( $P$ ) jo ek nonzero pattern ( $y$ ) permit karte hain. Unke beech, sirf straight column possible hai.

PICTURE. Ek staircase. $n = 1$ par ek half-sine fit hoti hai (single hump). $n = 2$ par ek full sine fit hoti hai (hump upar phir neeche). $n = 3$ par, dedh waves. Har ek ko ek higher load $P$ chahiye, $n^{2}$ ki tarah upar jaata hua.

$P_{n} = \frac{mode number squared n ^{2} π ^{2} E I}{L ^{2}}$

$n = 1$ :: ek half-wave, sabse sasta.
$n = 2, 3, \dots$ :: tighter waves, lekin har ek ko zyada load chahiye ( $n^{2}$ ).

Step 8 — Nature sabse sasta door leta hai

KYA. Jab aap slowly $P$ ko zero se badhate ho, pehle kaunsa special load milta hai? Sabse chhota, $n = 1$ : $P_{cr} = \frac{π ^{2} E I}{L ^{2}}, y (x) = A sin (\frac{π x}{L})$

KYU. Aap $n = 1$ ko skip karke $n = 2$ tak nahi pahunch sakte — column pehla door khulte hi buckle kar jaata hai. Isliye higher modes real solutions hain lekin column unka intezaar nahi karta (jab tak aap ise brace nahi karte pehle mode ko forbid karne ke liye). Shape ek single half-sine hump hai.

Ek honest caveat. Amplitude $A$ abhi bhi unknown hai! Hamari theory ne load aur shape deliver ki, lekin yeh nahi bataya ki woh kitna bow karta hai. Small-deflection (linear) theory literally $A$ nahi bata sakti — iske liye aapko nonlinear post-buckling theory chahiye. Yeh maths ki ek feature hai, koi galti nahi.

PICTURE. Winning mode: ek clean half-sine, peak middle mein, dono ends zero par pinned. Saath mein critical load labelled.

Step 9 — Har end condition, har degenerate case

KYA. Sab columns dono ends par pinned nahi hote. Alag supports change karte hain ki kaunsi wave fit hoti hai, effective length $L_{e} = K L$ se captured (dekho Column End Conditions and Effective Length): $P_{cr} = \frac{π ^{2} E I}{( K L ) ^{2}}$

KYU / PICTURE. Har support ke liye buckled hump sirf ek sine wave ka alag piece hai. $L_{e}$ us equivalent full pin-pin half-wave ki length hai jo us shape ke andar chhupi hui hai.

Ends	$K$	Wave jo fit hoti hai	Strength
Pin–Pin	$1.0$	full $L$ par half sine	baseline
Fixed–Fixed	$0.5$	dono ends clamped	strongest ( $4 \times$ )
Fixed–Pin	$\approx 0.7$	intermediate	$\approx 2 \times$
Fixed–Free	$2.0$	sirf quarter wave	weakest ( $\frac{1}{4} \times$ )

Woh degenerate & limiting cases jinse aapko kabhi surprise nahi hona chahiye:

Common mistake Woh cases jo formula secretly cover karta hai

$P < P_{cr}$ (threshold se neeche): akela solution $A = 0$ hai, straight column. Koi bent shape balance nahi kar sakti — buckling simply shuru nahi hui.
$L \to 0$ (bahut chhota column): $P_{cr} = π^{2} E I / L^{2} \to \infty$ . Euler infinite buckling load predict karta hai — impossible! Isliye ek chhota column kabhi buckle nahi karta; woh pehle yield/crush ho jaata hai. Euler sirf slender columns ke liye sach hai. Yield vs Buckling — Failure Mode Selection ke zariye cross over karo.
$L \to \infty$ (bahut lamba): $P_{cr} \to 0$ . Itna lamba column lagbhag kuch bhi nahi pe buckle kar jaata hai. Slenderness dushman hai.
$I$ sabse kamzor axis pick karta hai: ek cross-section ke paas do $I$ values hoti hain (ek per bending direction). Column us axis ke baare mein buckle karta hai jiske paas sabse chhota $I$ hai — hamesha dono check karo.
Higher modes ( $n \geq 2$ ): real, lekin sirf tab reachable hain jab aap physically column ko brace karo $n = 1$ hump forbid karne ke liye (jaise mid-span support). Middle mein bracing $n = 2$ unlock karta hai aur load char guna kar deta hai.

Ek-picture summary

Upar sab kuch, ek canvas par: force $P$ jo offset $y$ par act karta hai ek moment banata hai (Step 2), beam curvature se jawab deta hai (Step 3), dono ek wave equation mein combine ho jaate hain (Step 4), pins use quantise karte hain (Steps 6–7), aur sabse sasti wave jeet jaati hai (Step 8).

Recall Feynman retelling — plain words mein kaho wापas

Ek straight column par neeche push karo. Jab tak woh dead straight hai, push seedha middle se nikalta hai aur kuch bend nahi hota. Lekin suppose yeh ek baal sideways drift kare — ab downward push thoda off-centre act kar raha hai, aur off-centre push cheezein twist karti hai. Woh twist column ko thoda aur bend karta hai, jo push ko aur off-centre le jaata hai, jo aur hard twist karta hai… ek runaway. Kya runaway jeet jaata hai yeh ek tug-of-war par depend karta hai: push bend karne ki koshish karta hai, beam ki stiffness $E I$ seedha karne ki koshish karti hai. Woh tug-of-war likhne se ek equation milti hai jo kehti hai "shape middle ki taraf wापas us proportion mein curve karti hai jitna door woh hai" — wahi rule jo ek wave follow karta hai. Isliye bent column ek sine wave honi chahiye. Ends par pins ise bulge nahi karne dete, isliye wave exactly zero par start aur khatam honi chahiye — aur sirf wo sine waves fit hoti hain jisme ends ke beech half-humps ka pura number ho. Un fits mein se har ek ko apna khas load chahiye, aur sabse chhota, ek single hump, woh load hai jis par column actually give karta hai: $P_{cr} = π^{2} E I / L^{2}$ . Lamba column, chhota load ( $1/ L^{2}$ ki tarah jaata hai); stiffer material ya mota cross-section, bada load. Kaheen bhi material ki strength nahi aayi — buckling stiffness aur shape ke baare mein hai, bas.

Recall Quick self-test

Buckled shape ek sine hone par kyun majboor hai, parabola nahi? ::: Kyunki governing equation $y^{''} = - k^{2} y$ demand karti hai ki curvature shape ki minus ho, aur sirf sine/cosine mein yeh property hoti hai. Sabse chhota load ( $n = 1$ ) kyun jeet jaata hai? ::: Aap $P$ zero se badhate ho aur pehle door khulte hi buckle karte ho; higher modes ko zyada load chahiye. Ek chhota column $P_{cr} = π^{2} E I / L^{2}$ kyun follow nahi karta? ::: Jab $L \to 0$ toh yeh infinite load predict karta hai; column bahut pehle yield/crush ho jaata hai, isliye ek strength limit le leti hai. Theory hume kya batane mein fail hoti hai? ::: Bow ka amplitude $A$ — linear theory load aur shape deti hai lekin yeh nahi ki woh kitna bend karta hai.