2.2.18 · Physics › Fluid Mechanics
Intuition Ek-sentence idea
Navier–Stokes equation basically ==F = m a hai jo fluid ke ek chote se blob ke liye likhi gayi hai==, jahan forces hain pressure , viscous friction , aur gravity , aur acceleration ko moving fluid ke saath track kiya jaata hai (material derivative).
F = m a kyun nahi use karte?
Newton ka law particles ke baare mein hai. Fluid ek continuum hai infinitely many particles ka jo flow karte hain, deform karte hain, aur ek doosre se slide karte hain. Hum har molecule ko track nahi kar sakte. Isliye hum F = m a apply karte hain ek infinitesimal fluid element (control volume of mass ρ d V ) par aur poochte hain: iska acceleration kya hai, aur iske upar kaun-kaun si forces act karti hain? Woh answer, per unit volume mein likha hua, hi Navier–Stokes hai.
Definition Material (substantial) derivative
Velocity field u ( r , t ) hai. Ek fluid particle ki velocity depend karti hai dono time aur us position par jo usne move karke haasil ki hai. Uska sahi acceleration hai:
D t D u = ∂ t ∂ u + ( u ⋅ ∇ ) u
Pehla term hai local (unsteady) acceleration ; doosra hai convective acceleration (aap sirf ek faster region mein move karke speed up ho jaate ho).
Intuition Convective term kyun exist karta hai
Ek river mein khade ho jo narrow hoti hai. Bhale hi flow steady ho (∂ u / ∂ t = 0 ), ek baha hua patta speed up hoga jab woh narrow part mein jaayega — kyunki woh faster paani mein carry ho raha hai. Aapke fixed point par speed constant hai, lekin particle ki speed change hoti hai. Woh change ( u ⋅ ∇ ) u hai.
KAISE derive karte hain (chain rule, first principles):
Ek particle r ( t ) par hai jiske saath velocity u ( r ( t ) , t ) hai. Total derivative:
d t d u = ∂ t ∂ u + ∂ x ∂ u d t d x + ∂ y ∂ u d t d y + ∂ z ∂ u d t d z
Lekin d t d x = u x , etc., toh last teen terms exactly ( u ⋅ ∇ ) u hain. Ho gaya.
Ek fluid ka cube lo, volume d V = d x d y d z , mass d m = ρ d V . Teen types ki forces:
d F body = ρ g d V
gradient kyun deta hai, sirf ek value nahi
Pressure har face par andar ki taraf push karta hai. Agar pressure uniform hota, toh left face par push right face ke push ko cancel kar deta — net zero. Net force ke liye cube mein pressure ka difference chahiye, yaani ek gradient.
KAISE (derivation): Left x -face par force = P ( x ) d y d z (+ x direction mein). Right face par = − P ( x + d x ) d y d z (− x direction mein). Net:
d F x = [ P ( x ) − P ( x + d x )] d y d z = − ∂ x ∂ P d x d y d z
3D mein:
d F pressure = − ∇ P d V
Intuition Viscosity = fluid layers ke beech friction kyun hai
Agar paas-paas ke fluid layers alag-alag speed se move kar rahe hain, toh molecules ek doosre ko drag karte hain — momentum sideways diffuse hota hai. Newton's law of viscosity: shear stress τ = μ ∂ y ∂ u . Ek blob par net viscous force iske across stress mein difference se aati hai — ek spatial derivative ka aur derivative, isliye ek second derivative.
KAISE (incompressible Newtonian fluid ke liye derivation):
Ek face par stress = μ ∂ u x / ∂ y . Dono y -faces se x mein net force:
[ μ ∂ y ∂ u x y + d y − μ ∂ y ∂ u x y ] d x d z = μ ∂ y 2 ∂ 2 u x d V
x , y , z faces ke contributions ko sum karte hue (aur extra terms drop karne ke liye incompressibility ke liye ∇ ⋅ u = 0 use karte hue):
d F visc = μ ∇ 2 u d V
d V cancel karo aur D u / D t expand karo:
Intuition Ise ek sentence ki tarah padho
(mass density × blob ka acceleration) = (pressure push) + (viscous drag) + (weight). Har term literally un teen forces mein se ek hai jo humne build kiye.
Worked example Euler's equation recover karo
Forecast: ek ideal fluid mein viscosity nahi hoti. μ = 0 set karo.
Verify: ρ D t D u = − ∇ P + ρ g — exactly Euler's equation . ✓ Yeh step kyun? Ise inviscid case par reduce hona chahiye, jo humara known limit hai.
Worked example Hydrostatics recover karo
Forecast: fluid at rest, u = 0 .
Verify: 0 = − ∇ P + ρ g ⟹ ∇ P = ρ g ⟹ d z d P = − ρ g . Yeh hai P = P 0 + ρ g h (h neeche ki taraf measure kiya hua). ✓ Yeh step kyun? Koi motion nahi ⟹ pressure ko gravity balance karna chahiye, jo humara sabse basic fluid result hai.
Worked example Plane Poiseuille flow (plates ke beech steady flow)
Flow u x ( y ) plates ke beech y = 0 , h par, pressure gradient ∂ P / ∂ x = − G se driven. Steady, koi y , z velocity nahi. NS x -component reduce hota hai:
0 = G + μ d y 2 d 2 u x ⇒ d y 2 d 2 u x = − μ G
u x ( 0 ) = u x ( h ) = 0 (no-slip) ke saath do baar integrate karo:
u x ( y ) = 2 μ G y ( h − y )
Yeh step kyun? Parabola viscous pressure-driven flow ka fingerprint hai — walls par zero, middle mein max.
Common mistake "Acceleration bas
∂ u / ∂ t hai."
Kyun sahi lagta hai: particle mechanics mein, a = d v / d t ek point par aur kaam ho jaata hai.
Flaw: ek field ke liye, particle ek naye jagah move karta hai jahan u alag hota hai. Tumhe ( u ⋅ ∇ ) u add karna hoga.
Fix: "fluid ka acceleration" ke liye hamesha material derivative D / D t use karo.
Common mistake "Pressure khud flow drive karta hai, toh term
− P hona chahiye."
Kyun sahi lagta hai: bada pressure = bada push.
Flaw: uniform pressure saari sides se equally push karta hai → zero net force. Sirf differences matter karte hain.
Fix: force per volume − ∇ P hai, gradient.
Common mistake "Viscous term
μ ∇ u hona chahiye (first derivative)."
Kyun sahi lagta hai: Newton's law of viscosity mein ek first derivative hai, τ = μ ∂ u / ∂ y .
Flaw: τ ek stress hai; net force blob ke across stress mein difference hai → ek aur derivative.
Fix: viscous force per volume = μ ∇ 2 u .
Recall Feynman: ek 12-saal ke bachche ko samjhao
Socho paani ka ek chota sa floating cube. Teen cheezein ise push karti hain: uske aas-paas ka paani ek side se doosri side se zyada squeeze kar raha hai (pressure), paas-paas ka paani slide karke ise rub kar raha hai (stickiness/viscosity), aur gravity ise neeche kheench rahi hai. Newton ne kaha push = mass × speed-kitni-tezi-se-badal-rahi-hai. Tricky part yeh hai: cube speed up hota hai dono isliye kyunki river change ho rahi hai aur isliye kyunki cube ek faster jagah float karta hai. Saari pushes add karo, unhe mass times speed-change ke barabar set karo, aur woh poora sentence hi Navier–Stokes equation hai.
Mnemonic Structure yaad karo
"My Pal Viscous Goes" → M ass·accel = P ressure + V iscous + G ravity.
Acceleration = "Local + Carried" (∂/∂t + u ⋅ ∇ ).
Material derivative likho aur dono terms ko naam do.
Pressure force − ∇ P kyun hai aur − P kyun nahi?
Viscosity ke liye second derivative kyun?
NS kya ban jaata hai agar μ = 0 ho? Agar u = 0 ho?
Material derivative D u / D t kya hota hai? ∂ u / ∂ t + ( u ⋅ ∇ ) u — local (unsteady) plus convective acceleration.
Convective term ( u ⋅ ∇ ) u kyun appear karta hai? Ek fluid particle velocity change karta hai different velocity ke region mein move karke, steady flow mein bhi.
Fluid element par net pressure force per unit volume? − ∇ P (sirf pressure differences net force dete hain).
Pressure force gradient kyun hai, sirf P kyun nahi? Uniform pressure saare faces par equally push karta hai → cancel ho jaata hai; sirf element ke across difference net force deta hai.
Viscous force per unit volume (incompressible Newtonian)? Viscosity second spatial derivative kyun hai? Stress μ ∂ u / ∂ y hai; net force blob ke across stress mein difference hai → ek aur derivative.
Full incompressible Navier–Stokes equation? ρ ( ∂ t u + ( u ⋅ ∇ ) u ) = − ∇ P + μ ∇ 2 u + ρ g .
Incompressible NS ke saath kaun sa constraint aata hai? NS with μ = 0 kya ban jaata hai? Euler's equation:
ρ D u / D t = − ∇ P + ρ g .
NS with u = 0 kya ban jaata hai? Hydrostatics:
∇ P = ρ g , yaani
d P / d z = − ρ g .
Plane Poiseuille flow ka velocity profile? u x ( y ) = 2 μ G y ( h − y ) , ek parabola (walls par no-slip).
NS kis physical principle se derive hoti hai? Newton's second law
F = m a jo ek infinitesimal fluid element par per unit volume apply ki gayi hai.
Infinitesimal element mass rho dV
Material derivative Du/Dt
Chain rule on u of r and t