3.6.4 · D1 · Physics › Spacecraft Structures & Systems Engineering › Hooke's law in 3D — generalized stress-strain (tensor)
Jab tum kisi solid ko push karte ho ya twist karte ho, woh deform hota hai — aur ek direction mein push karne se uska size har direction mein quietly change hota hai. "Saari pushes → saari deformations" ko capture karne ke liye humein ek aisa bookkeeping device chahiye jisme har direction-pair ke liye enough slots hon: woh device hai ek tensor , aur poora 3D Hooke's law bas ek bada linear rule hai jo "push tensor" ko "deform tensor" se connect karta hai.
Yeh page assume karta hai ki tumne kuch nahi dekha. Hum har letter, index, aur picture ko build karenge jo parent note pe depend karta hai, ek aisi order mein jahan har idea sirf usse pehle waale idea pe rest karta hai.
Kisi bhi physics se pehle, humein ek tarika chahiye yeh kehne ka ki "yeh direction, woh nahi".
Definition Teen axes aur unke labels
Hum teen mutually perpendicular directions choose karte hain aur unhe x , y , z kehte hain — ya, jab hum unhe count karna chahte hain, unhe ==directions 1 , 2 , 3 == kehte hain. Yeh wahi teen arrows hain; 1 = x , 2 = y , 3 = z . Unhe number karne se hum formulas mein x, y, z har baar spell out karne ki jagah ek counter use kar sakte hain.
Letters ki jagah numbers kyun? Kyunki jaldi hi humein aise quantities milenge jo do directions ek saath carry karte hain (ek face + ek force). σ 12 likhna kaafi cleaner hai "the-thing-on-the-x-face-pointing-in-y" se. Upar ki picture puri coordinate stage hai — is topic mein har arrow, angle aur tensor slot inhi teen axes mein se kisi ek ki taraf point karta hai.
Definition Subscript ka matlab
Subscript (ek letter ke neeche ka chota number) bas ek address hai. u i matlab hai "u ka i -wa piece": u 1 hai x -piece, u 2 hai y -piece, u 3 hai z -piece. Letter i ek placeholder hai jo "1, 2, 3 mein se jo bhi tumhe pasand ho" ke liye khada hai.
Intuition Ek index vs do indices — picture
Ek index (v i ) = ek arrow. Ise describe karne ke liye ek direction chahiye: x , y , z ke along kitna. Teen numbers.
Do indices (M ij ) = numbers ki ek table (grid). Row i , column j . Nau numbers. Tum do indices tab use karte ho jab ek quantity ko sense banane ke liye do directions chahiye hon — jaise "direction j mein force jo direction i face karne wale face across act kar raha hai".
M ij ko zor se padho "M-row-i-column-j". Yeh aadat akele poore topic ko unlock kar deti hai.
Vector ek arrow hai jisme length aur direction hoti hai. Humare axes mein yeh teen numbers ( v 1 , v 2 , v 3 ) ke roop mein store hota hai, ek per axis. Displacement u (ek point kitna move hua) ek vector hai; force ek vector hai.
Bold u matlab hai "poora arrow", jabki u i matlab hai "uska ek chosen piece". Same object, do zoom levels.
Definition Tensor (rank = indices ki sankhya)
Tensor ek general naam hai "numbers ka ek box, jisme har ek directions se labelled hai".
Rank 0 = ek plain number (ek scalar ), koi index nahi. Example: temperature.
Rank 1 = ek vector, ek index, 3 numbers.
Rank 2 = ek grid, do indices, 3 × 3 = 9 numbers.
Rank 4 = char indices, 3 × 3 × 3 × 3 = 81 numbers.
Topic ko rank-2 tensor kyun chahiye? Kyunki dono stress aur strain deformation states describe karte hain jinhe pin down karne ke liye do directions chahiye. Rank-4 kyun? Kyunki ek 3 × 3 grid ko doosre 3 × 3 grid se, slot by slot, relate karne ke liye ek aisa rule chahiye jisme chaar address slots hon — yahi stiffness tensor C ij k l hai.
Stress σ ij force hai jo ek area pe spread hai, pascals (Pa = N / m 2 ) mein measure hota hai. Do indices yeh read karte hain: pehla index i = force kis face pe act kar raha hai (woh face jiska outward normal axis i ke along point karta hai); doosra index j = force khud kis direction mein point karta hai.
σ 11 : x ke along force, x face karne wale face pe → face ke seedha andar pull/push (normal stress).
σ 12 : y ke along force, x face karne wale face pe → face ke along sideways drag (shear stress, aksar τ 12 likha jaata hai).
Intuition Normal vs shear — picture
Agar chota arrow apne face ke perpendicular hai, toh yeh stretch ya squash karta hai → normal stress (diagonal entries σ 11 , σ 22 , σ 33 ). Agar arrow apne face ke flat along hai, toh yeh layers ko ek doosre ke past slide karta hai → shear stress (off-diagonal entries).
Greek letter hai sigma σ — yahan iska matlab hamesha stress hai. Symbol τ (tau ) sirf off-diagonal shear entries ka nickname hai.
Definition Sign convention (ise ek baar yaad kar lo)
Normal stress: tension positive hai , compression negative. Toh ek rod jo pull ho rahi hai uska σ 11 > 0 ; squeeze ho rahi rod ka σ 11 < 0 . Inward push karne wala pressure p deta hai σ 11 = σ 22 = σ 33 = − p .
Normal strain: stretch positive hai , shrink negative. Rod ko pull karo aur ε 11 > 0 ; Poisson sideways shrink deta hai ε 22 , ε 33 < 0 .
Shear sign: σ ij (jab i = j ) positive hota hai jab, us face pe jiska outward normal + i ke along point karta hai, force + j ke along point kare. Face ya force direction dono mein se kisi ek ko flip karo aur sign flip ho jaata hai. Shear strain ε ij positive hoti hai jab originally-perpendicular + i aur + j material lines ke beech ka angle kam hota hai.
Parent note mein har worked example exactly isi "tension +, compression −" convention use karta hai.
Strain ε ij measure karta hai ki material kitna deform hua, ek fraction ke roop mein (stretch ÷ original length). Iske koi units nahi hain. Normal strain ε 11 = "x ke along length mein fractional change"; shear strain ε 12 = "ek right angle kitna skewed hua".
Fractional kyun, absolute nahi? 2 mm wire ke liye 1 mm stretch bahut zyada hai aur 2 m beam ke liye kuch nahi. Original length se divide karne se strain deformation ki ek property ban jaati hai, object ke size se independent — exactly wahi jo ek material law ko chahiye. Greek letter hai epsilon ε ; iska matlab hamesha strain hai.
Definition Shear strain ke do flavours
Tensor shear ε 12 : neat symmetric grid mein jaata hai.
Engineering shear γ 12 = 2 ε 12 (gamma): woh total angle change jo tum protractor se measure karte.
Yeh factor of 2 se differ karte hain. Parent note ka "factor-of-2 trap" puri tarah se inhe straight rakhne ke baare mein hai.
Common mistake "Small strain" sirf linearized measure hai
Definition ε ij = 2 1 ( ∂ j u i + ∂ i u j ) infinitesimal (small-strain) tensor hai: yeh displacement gradient mein sirf linear terms rakhta hai aur quadratic ones ko throw away kar deta hai.
Yeh safe kyun lagta hai: stiff metals aur typical spacecraft loads ke liye strains tiny hoti hain (∼ 1 0 − 3 ), toh dropped quadratic terms negligible hain aur yeh exact enough hai.
Yahan toot ta hai: large deformations ke liye (rubber seals, inflatables, buckled panels jo tens of percent se stretch hain) discarded quadratic part matter karta hai aur tumhe ek finite-strain measure pe switch karna hoga (jaise Green–Lagrange strain). Is page pe aur parent note mein sab kuch small strain assume karta hai.
Intuition Grid apne diagonal ke across mirror kyun karta hai
Agar σ 12 (drag on the x -face pointing y ) equal nahi hota σ 21 (drag on the y -face pointing x ) se, toh chota cube ek net twist feel karta aur forever spin karta rehta — rest mein body ke liye impossible. Moment (torque) balance force karta hai σ ij = σ j i . Wahi argument strain ko bhi symmetric banata hai.
Is mirror ki wajah se, 9 slots mein sirf 6 independent numbers hain: 3 diagonal, plus 3 off-diagonal pairs. Yahi "6" hai isiliye engineers ek tensor ko ek 6-slot list mein pack kar sakte hain (Voigt notation).
Nau numbers σ ij teen coordinate faces pe defined hain. Lekin ek real crack ya bolt material ko kisi bhi tilted plane ke along cut kar sakta hai. Woh arbitrary plane kaisi force feel karta hai?
Definition Traction vector aur Cauchy's relation
Material ko ek aisi plane se slice karo jiska outward normal unit arrow n = ( n 1 , n 2 , n 3 ) hai. Woh force per area jo us cut ke across transmit hoti hai woh traction vector t hai. Iske pieces stress tensor se build hote hain:
t i = σ ij n j ( sum over j ) .
Yeh hai Cauchy's stress theorem : stress tensor exactly woh machine hai jo "cut kis taraf face kar rahi hai" (n ) ko "woh kaisi force feel karti hai" (t ) mein turn karta hai. Dekho Stress tensor and Cauchy's relation .
Yahan yeh kyun matter karta hai: yeh batata hai kyun stress ko do indices chahiye. Ek index (j , summed) plane ki orientation read karta hai; free index (i ) resulting force direction deliver karta hai. Toh σ ij sirf "teen faces pe nau numbers" nahi hai — yeh us point ke through har imaginable slice pe force ka complete rule hai.
Definition Kronecker delta
δ ij = { 1 0 if i = j if i = j
Yeh ek chota "kya yeh do directions same hain?" switch hai: on (1 ) jab indices match hote hain, off (0 ) otherwise. Ek grid ke roop mein yeh identity matrix hai (diagonal pe ones, baaki zeros).
Topic ko yeh kyun chahiye: yeh ek ek direction-blind building block hai jo available hai. Jab ek material har direction mein same behave karta hai (isotropic), toh aap sirf δ 's se bani tensors se uska law build kar sakte ho — isi tarah parent note 81 constants se 2 tak pahunchta hai.
Definition Repeated index ⇒ add it up
Jab ek index ek term mein do baar appear karta hai, tum secretly 1 , 2 , 3 pe sum karte ho. Toh
C_{ijkl}\,\varepsilon_{kl} \;\equiv\; \sum_{k=1}^{3}\sum_{l=1}^{3} C_{ijkl}\,\varepsilon_{kl}.$$
ε k k ke peeche ki picture
Teen diagonal strains ko add karna = "x ke along stretch" + "y ke along" + "z ke along" add karna = volume mein total fractional change . Isiliye ε k k ko volumetric strain kehte hain: yeh ek swelling/shrinking meter hai.
Common mistake Free index vs summed index
Galti: sochna ki har repeated letter ka matlab "sum" hai.
Fix: ek index jo ek term mein ek baar appear karta hai (ek free index, jaise σ ij = C ij k l ε k l mein i , j ) summed nahi hota — yeh label karta hai ki tum kaun si equation dekh rahe ho. Sirf woh index summed hota hai jo do baar appear karta hai (yahan k aur l ). Free indices equation ke dono sides pe match hone chahiye.
Definition Partial derivative aur uska shorthand
∂ x j ∂ u i puchta hai: "jab main direction x j mein ek tiny bit step karta hoon, displacement-piece u i kitni tezi se change hota hai?" Yeh ek slope hai — rise (u i ) over run (x j ) — measure kiya jaata hai jab doosre directions fixed rakhe jaate hain. Hum ise abbreviate karte hain:
∂ j u i ≡ ∂ x j ∂ u i .
Chota symbol ∂ j bas matlab hai "axis j ke along move karte waqt change ki rate" — kuch nahi.
Yeh tool kyun, plain fraction nahi? Ek plain ratio Δ u /Δ x depend karta hai ki tum kitna bada step lete ho. Derivative woh limit hai jab step zero tak shrink ho jaata hai, ek single point pe local stretch rate deta hai — exactly wahi jo strain ko chahiye, kyunki strain ek bent structure ke andar point to point vary kar sakti hai. Nau slopes ∂ j u i displacement-gradient grid form karte hain; uska symmetric half strain tensor hai (dekho Strain tensor and displacement gradient ).
Intuition Symmetric vs antisymmetric half — stretch vs spin
Koi bhi grid ek mirror-symmetric part aur ek anti-mirror part mein split hota hai. Symmetric part genuine stretching/shearing hai (energy store karta hai). Antisymmetric part pure rotation hai — poora cube sirf rotate karta hai bina shape change kiye, koi energy store nahi karta. Strain sirf symmetric half rakhti hai, isiliye ε ij = 2 1 ( ∂ j u i + ∂ i u j ) .
Ab jab stress, strain, indices, aur summation sab build ho gaye, hum parent note ka central rule uski apni definition ke roop mein state kar sakte hain.
C ij k l 81 se 21 se 2 kyun shrink hota hai — uski symmetries
C ij k l 3 4 = 81 slots se shuru hota hai, lekin teen symmetries ise slash karti hain:
Minor symmetry (stress side): kyunki σ ij = σ j i , i ↔ j swap karna matter nahi kar sakta, toh C ij k l = C j ik l .
Minor symmetry (strain side): kyunki ε k l = ε l k , k ↔ l swap karna matter nahi kar sakta, toh C ij k l = C ij l k .
Major symmetry (energy): stored elastic energy ek single number hai, toh poore pair ij ↔ k l ko swap karne se C unchanged rehta hai: C ij k l = C k l ij .
Do minor symmetries 81 → 36 kar deti hain; major symmetry 36 → 21 constants kar deti hai fully anisotropic solid ke liye. Dekho Isotropic vs anisotropic materials (composites) .
Definition Young's modulus
E aur Poisson's ratio ν
==Young's modulus E == (units: Pa): stiffness. Bada E = stretch karna mushkil (steel); chota E = aasaan (rubber). Yeh stress-vs-strain ka slope hai jab tum ek taraf pull karte ho.
==Poisson's ratio ν == (nu, koi units nahi): "bulge" dial. Jab tum x ke along kisi fraction se stretch karte ho, sideways shrink utni hi badi hoti hai jitni ν times. Dekho Poisson's ratio and material limits .
Hooke 3D sigma equals C epsilon
Khud test karo — right side cover karo aur reveal karne se pehle answer do.
u i mein subscript i kya pick out karta hai?Vector u ka i -wa direction-piece (u 1 = x -part, etc.).
Ek symmetric 3 × 3 tensor mein kitne independent numbers hote hain, aur kyun? 6 — teen diagonal plus teen off-diagonal pairs, kyunki grid mirror karta hai (σ ij = σ j i ).
σ 12 mein do indices ka kya matlab hai?Direction 2 (y ) mein point karta force, us face pe act karta hai jiska normal direction 1 (x ) mein point karta hai — ek shear stress.
Normal stress aur strain ka sign convention kya hai? Tension/stretch positive, compression/shrink negative; ek pressure p deta hai σ ii = − p .
Traction relation t i = σ ij n j tumhe kya batata hai? Stress tensor plane ke normal n ko force-per-area t mein turn karta hai jo woh plane feel karta hai (Cauchy's theorem).
δ ij i = j hone pe aur i = j hone pe kya equal hai?1 jab i = j , 0 jab i = j (identity switch).
ε k k ko poora likh ke batao ki physically iska kya matlab hai.ε 11 + ε 22 + ε 33 — volume mein fractional change (volumetric strain).
C ij k l ε k l mein kaun sa index summed hai aur kaun sa free?k aur l summed hain (har ek repeated hai); i aur j free hain (equation label karte hain).
Strain displacement-gradient ∂ j u i ka sirf symmetric half kyun rakhti hai? Antisymmetric half pure rotation hai, jo koi shape change nahi karta aur koi elastic energy store nahi karta.
∂ j kiska shorthand hai?∂ / ∂ x j ka shorthand, axis j ke along step karte waqt change ki rate.
Small-strain formula ε ij = 2 1 ( ∂ j u i + ∂ i u j ) kab break down karta hai? Large deformations ke liye — yeh quadratic gradient terms drop kar deta hai, toh tab finite-strain measures zaroori hain.
Fully anisotropic solid ke liye kitne independent constants hain, aur isotropy ke baad? 21 (81 se minor + major symmetries ke through), isotropy ke liye 2 tak gir jaata hai.
Engineering shear γ 12 aur tensor shear ε 12 ke beech kya relation hai? γ 12 = 2 ε 12 .
Strain ke liye hum ratio Δ u /Δ x ki jagah partial derivative kyun use karte hain? Derivative ek point pe local, step-size-independent stretch rate hai; ek ratio depend karta hai ki tum kitna bada step lete ho.
Ek isotropic solid ko kitne material constants chahiye? Do (jaise E aur ν , ya λ aur μ ).