Intuition The one-sentence idea
The Navier–Stokes equation is just ==F ⃗ = m a ⃗ \vec F = m\vec a F = m a written for a tiny blob of fluid==, where the forces are pressure , viscous friction , and gravity , and the acceleration is tracked following the moving fluid (the material derivative).
Intuition Why not just use
F ⃗ = m a ⃗ \vec F = m\vec a F = m a directly?
Newton's law is about particles . A fluid is a continuum of infinitely many particles that flow, deform, and slide past each other. We can't track each molecule. So we apply F ⃗ = m a ⃗ \vec F = m\vec a F = m a to an infinitesimal fluid element (a control volume of mass ρ d V \rho\,dV ρ d V ) and ask: what is its acceleration, and what forces act on it? The answer, written per unit volume, is Navier–Stokes.
Definition Material (substantial) derivative
The velocity field is u ⃗ ( r ⃗ , t ) \vec u(\vec r, t) u ( r , t ) . A fluid particle's velocity depends on both time and the position it has moved to. Its true acceleration is
D u ⃗ D t = ∂ u ⃗ ∂ t + ( u ⃗ ⋅ ∇ ) u ⃗ \frac{D\vec u}{Dt} = \frac{\partial \vec u}{\partial t} + (\vec u\cdot\nabla)\vec u D t D u = ∂ t ∂ u + ( u ⋅ ∇ ) u
The first term is local (unsteady) acceleration ; the second is convective acceleration (you speed up just by moving into a faster region).
Intuition WHY the convective term exists
Stand in a river that narrows. Even if the flow is steady (∂ u ⃗ / ∂ t = 0 \partial\vec u/\partial t = 0 ∂ u / ∂ t = 0 ), a drifting leaf speeds up as it enters the narrow part — because it is carried into faster water. The speed at your fixed point is constant, but the particle's speed changes. That change is ( u ⃗ ⋅ ∇ ) u ⃗ (\vec u\cdot\nabla)\vec u ( u ⋅ ∇ ) u .
HOW to derive it (chain rule, first principles):
A particle is at r ⃗ ( t ) \vec r(t) r ( t ) with velocity u ⃗ ( r ⃗ ( t ) , t ) \vec u(\vec r(t), t) u ( r ( t ) , t ) . Total derivative:
d u ⃗ d t = ∂ u ⃗ ∂ t + ∂ u ⃗ ∂ x d x d t + ∂ u ⃗ ∂ y d y d t + ∂ u ⃗ ∂ z d z d t \frac{d\vec u}{dt} = \frac{\partial\vec u}{\partial t} + \frac{\partial \vec u}{\partial x}\frac{dx}{dt}+\frac{\partial\vec u}{\partial y}\frac{dy}{dt}+\frac{\partial\vec u}{\partial z}\frac{dz}{dt} d t d u = ∂ t ∂ u + ∂ x ∂ u d t d x + ∂ y ∂ u d t d y + ∂ z ∂ u d t d z
But d x d t = u x \frac{dx}{dt}=u_x d t d x = u x , etc., so the last three terms are exactly ( u ⃗ ⋅ ∇ ) u ⃗ (\vec u\cdot\nabla)\vec u ( u ⋅ ∇ ) u . Done.
Take a cube of fluid, volume d V = d x d y d z dV = dx\,dy\,dz d V = d x d y d z , mass d m = ρ d V dm=\rho\,dV d m = ρ d V . Three force types:
d F ⃗ body = ρ g ⃗ d V d\vec F_{\text{body}} = \rho\,\vec g\,dV d F body = ρ g d V
Intuition WHY pressure gives a
gradient , not just a value
Pressure pushes inward on every face. If pressure were uniform, the push on the left face cancels the push on the right — net zero. A net force needs a difference in pressure across the cube, i.e. a gradient.
HOW (derivation): Force on left x x x -face = P ( x ) d y d z = P(x)\,dy\,dz = P ( x ) d y d z (pointing + x +x + x ). On right face = − P ( x + d x ) d y d z = -P(x+dx)\,dy\,dz = − P ( x + d x ) d y d z (pointing − x -x − x ). Net:
d F x = [ P ( x ) − P ( x + d x ) ] d y d z = − ∂ P ∂ x d x d y d z dF_x = [P(x) - P(x+dx)]\,dy\,dz = -\frac{\partial P}{\partial x}dx\,dy\,dz d F x = [ P ( x ) − P ( x + d x )] d y d z = − ∂ x ∂ P d x d y d z
In 3D:
d F ⃗ pressure = − ∇ P d V d\vec F_{\text{pressure}} = -\nabla P\,dV d F pressure = − ∇ P d V
Intuition WHY viscosity = friction between fluid layers
If neighbouring fluid layers move at different speeds, molecules drag each other — momentum diffuses sideways. Newton's law of viscosity: shear stress τ = μ ∂ u ∂ y \tau = \mu\,\frac{\partial u}{\partial y} τ = μ ∂ y ∂ u . The net viscous force on a blob comes from the difference in stress across it — another spatial derivative of a derivative, hence a second derivative.
HOW (derivation for incompressible Newtonian fluid):
Stress on a face = μ ∂ u x / ∂ y = \mu\,\partial u_x/\partial y = μ ∂ u x / ∂ y . Net force in x x x from the two y y y -faces:
[ μ ∂ u x ∂ y ∣ y + d y − μ ∂ u x ∂ y ∣ y ] d x d z = μ ∂ 2 u x ∂ y 2 d V \left[\mu\frac{\partial u_x}{\partial y}\Big|_{y+dy} - \mu\frac{\partial u_x}{\partial y}\Big|_{y}\right]dx\,dz = \mu\frac{\partial^2 u_x}{\partial y^2}\,dV [ μ ∂ y ∂ u x y + d y − μ ∂ y ∂ u x y ] d x d z = μ ∂ y 2 ∂ 2 u x d V
Summing the contributions from x , y , z x,y,z x , y , z faces (and using ∇ ⋅ u ⃗ = 0 \nabla\cdot\vec u=0 ∇ ⋅ u = 0 for incompressibility to drop extra terms):
d F ⃗ visc = μ ∇ 2 u ⃗ d V d\vec F_{\text{visc}} = \mu\nabla^2\vec u\,\,dV d F visc = μ ∇ 2 u d V
Cancel d V dV d V and expand D u ⃗ / D t D\vec u/Dt D u / D t :
Intuition Read it like a sentence
(mass density × acceleration of the blob) = (pressure push) + (viscous drag) + (weight). Each term is literally one of the three forces we built.
Worked example Recover Euler's equation
Forecast: an ideal fluid has no viscosity. Set μ = 0 \mu=0 μ = 0 .
Verify: ρ D u ⃗ D t = − ∇ P + ρ g ⃗ \rho\frac{D\vec u}{Dt} = -\nabla P + \rho\vec g ρ D t D u = − ∇ P + ρ g — exactly Euler's equation . ✓ Why this step? It must reduce to the inviscid case, our known limit.
Worked example Recover hydrostatics
Forecast: fluid at rest, u ⃗ = 0 \vec u = 0 u = 0 .
Verify: 0 = − ∇ P + ρ g ⃗ 0 = -\nabla P + \rho\vec g 0 = − ∇ P + ρ g ⟹ ∇ P = ρ g ⃗ \nabla P = \rho\vec g ∇ P = ρ g ⟹ d P d z = − ρ g \frac{dP}{dz}=-\rho g d z d P = − ρ g . That's P = P 0 + ρ g h P = P_0 + \rho g h P = P 0 + ρ g h (with h h h measured downward). ✓ Why this step? No motion ⟹ pressure must balance gravity, our most basic fluid result.
Worked example Plane Poiseuille flow (steady flow between plates)
Flow u x ( y ) u_x(y) u x ( y ) between plates at y = 0 , h y=0,h y = 0 , h , driven by pressure gradient ∂ P / ∂ x = − G \partial P/\partial x = -G ∂ P / ∂ x = − G . Steady, no y , z y,z y , z velocity. NS x x x -component reduces to:
0 = G + μ d 2 u x d y 2 ⇒ d 2 u x d y 2 = − G μ 0 = G + \mu\frac{d^2 u_x}{dy^2}\;\Rightarrow\; \frac{d^2u_x}{dy^2}=-\frac{G}{\mu} 0 = G + μ d y 2 d 2 u x ⇒ d y 2 d 2 u x = − μ G
Integrate twice with u x ( 0 ) = u x ( h ) = 0 u_x(0)=u_x(h)=0 u x ( 0 ) = u x ( h ) = 0 (no-slip):
u x ( y ) = G 2 μ y ( h − y ) u_x(y) = \frac{G}{2\mu}\,y(h-y) u x ( y ) = 2 μ G y ( h − y )
Why this step? A parabola is the fingerprint of viscous pressure-driven flow — zero at walls, max in the middle.
Common mistake "Acceleration is just
∂ u ⃗ / ∂ t \partial\vec u/\partial t ∂ u / ∂ t ."
Why it feels right: in particle mechanics, a = d v / d t a = dv/dt a = d v / d t at a point and we're done.
The flaw: for a field , the particle moves to a new place where u ⃗ \vec u u differs. You must add ( u ⃗ ⋅ ∇ ) u ⃗ (\vec u\cdot\nabla)\vec u ( u ⋅ ∇ ) u .
Fix: always use the material derivative D / D t D/Dt D / D t for "acceleration of the fluid."
Common mistake "Pressure itself drives the flow, so the term should be
− P -P − P ."
Why it feels right: big pressure = big push.
The flaw: a uniform pressure pushes equally from all sides → zero net force. Only differences matter.
Fix: the force per volume is − ∇ P -\nabla P − ∇ P , the gradient.
Common mistake "Viscous term should be
μ ∇ u ⃗ \mu\nabla\vec u μ ∇ u (first derivative)."
Why it feels right: Newton's law of viscosity has a first derivative, τ = μ ∂ u / ∂ y \tau=\mu\,\partial u/\partial y τ = μ ∂ u / ∂ y .
The flaw: τ \tau τ is a stress ; the net force is the difference in stress across the blob → one more derivative.
Fix: viscous force per volume = μ ∇ 2 u ⃗ =\mu\nabla^2\vec u = μ ∇ 2 u .
Recall Feynman: explain to a 12-year-old
Imagine a tiny floating cube of water. Three things push it: the water around it squeezing harder on one side than the other (pressure), neighbouring water sliding and rubbing it (stickiness/viscosity), and gravity pulling it down. Newton said push = mass × how-fast-speed-changes. The tricky bit: the cube speeds up both because the river is changing and because the cube floats into a faster spot. Add all the pushes, set them equal to mass times speed-change, and that whole sentence is the Navier–Stokes equation.
Mnemonic Remember the structure
"My Pal Viscous Goes" → M ass·accel = P ressure + V iscous + G ravity.
Acceleration = "Local + Carried" (∂/∂t + u ⃗ ⋅ ∇ \vec u\cdot\nabla u ⋅ ∇ ).
Write the material derivative and name both terms.
Why is the pressure force − ∇ P -\nabla P − ∇ P and not − P -P − P ?
Why second derivative for viscosity?
What does NS reduce to if μ = 0 \mu=0 μ = 0 ? If u ⃗ = 0 \vec u=0 u = 0 ?
What is the material derivative D u ⃗ / D t D\vec u/Dt D u / D t ? ∂ u ⃗ / ∂ t + ( u ⃗ ⋅ ∇ ) u ⃗ \partial\vec u/\partial t + (\vec u\cdot\nabla)\vec u ∂ u / ∂ t + ( u ⋅ ∇ ) u — local (unsteady) plus convective acceleration.
Why does the convective term ( u ⃗ ⋅ ∇ ) u ⃗ (\vec u\cdot\nabla)\vec u ( u ⋅ ∇ ) u appear? A fluid particle changes velocity by moving into a region of different velocity, even in steady flow.
Net pressure force per unit volume on a fluid element? − ∇ P -\nabla P − ∇ P (only pressure
differences give net force).
Why is the pressure force a gradient, not just P P P ? Uniform pressure pushes equally on all faces → cancels; only the difference across the element yields net force.
Viscous force per unit volume (incompressible Newtonian)? μ ∇ 2 u ⃗ \mu\nabla^2\vec u μ ∇ 2 u .
Why is viscosity a second spatial derivative? Stress is
μ ∂ u / ∂ y \mu\,\partial u/\partial y μ ∂ u / ∂ y ; net force is the
difference in stress across the blob → another derivative.
Full incompressible Navier–Stokes equation? ρ ( ∂ t u ⃗ + ( u ⃗ ⋅ ∇ ) u ⃗ ) = − ∇ P + μ ∇ 2 u ⃗ + ρ g ⃗ \rho(\partial_t\vec u + (\vec u\cdot\nabla)\vec u) = -\nabla P + \mu\nabla^2\vec u + \rho\vec g ρ ( ∂ t u + ( u ⋅ ∇ ) u ) = − ∇ P + μ ∇ 2 u + ρ g .
What constraint accompanies incompressible NS? Continuity:
∇ ⋅ u ⃗ = 0 \nabla\cdot\vec u = 0 ∇ ⋅ u = 0 .
NS with μ = 0 \mu=0 μ = 0 reduces to? Euler's equation:
ρ D u ⃗ / D t = − ∇ P + ρ g ⃗ \rho\,D\vec u/Dt = -\nabla P + \rho\vec g ρ D u / D t = − ∇ P + ρ g .
NS with u ⃗ = 0 \vec u=0 u = 0 reduces to? Hydrostatics:
∇ P = ρ g ⃗ \nabla P = \rho\vec g ∇ P = ρ g , i.e.
d P / d z = − ρ g dP/dz=-\rho g d P / d z = − ρ g .
Velocity profile for plane Poiseuille flow? u x ( y ) = G 2 μ y ( h − y ) u_x(y)=\frac{G}{2\mu}y(h-y) u x ( y ) = 2 μ G y ( h − y ) , a parabola (no-slip at walls).
What physical principle is NS derived from? Newton's second law
F ⃗ = m a ⃗ \vec F=m\vec a F = m a applied to an infinitesimal fluid element, per unit volume.
Infinitesimal element mass rho dV
Material derivative Du/Dt
Chain rule on u of r and t
Intuition Hinglish mein samjho
Dekho, Navier–Stokes equation koi alag-thalag formula nahi hai — yeh sirf Newton ka F ⃗ = m a ⃗ \vec F = m\vec a F = m a hai, but ek chhote se paani ke tukde (fluid element) ke liye likha gaya. Mass density ρ \rho ρ guna acceleration, equal to total force per volume. Force kahan se? Teen jagah se: pressure ka difference (isliye − ∇ P -\nabla P − ∇ P , kyunki agar pressure har taraf barabar ho toh net push zero), viscosity yaani layers ke beech ki ragad (μ ∇ 2 u ⃗ \mu\nabla^2\vec u μ ∇ 2 u ), aur gravity (ρ g ⃗ \rho\vec g ρ g ).
Sabse important twist hai acceleration ka. Particle mechanics mein hum d v / d t dv/dt d v / d t likhte hain, but fluid mein particle khud move karke nayi jagah pahunch jata hai jahan velocity alag hai. Isliye acceleration ke do part hote hain: ∂ u ⃗ / ∂ t \partial\vec u/\partial t ∂ u / ∂ t (time ke saath change) aur ( u ⃗ ⋅ ∇ ) u ⃗ (\vec u\cdot\nabla)\vec u ( u ⋅ ∇ ) u (move karke fast region mein ghus jaane wala change). Isko material derivative D u ⃗ / D t D\vec u/Dt D u / D t bolte hain. River ka example yaad rakho — narrow part mein leaf tez ho jata hai bina time ke flow change hue.
Viscous term mein ek common galti: log μ ∇ u ⃗ \mu\nabla\vec u μ ∇ u likh dete hain kyunki Newton's viscosity law mein single derivative hota hai (τ = μ ∂ u / ∂ y \tau=\mu\,\partial u/\partial y τ = μ ∂ u / ∂ y ). But τ \tau τ toh stress hai — net force toh stress ke difference se aati hai, toh ek aur derivative lagta hai, isliye μ ∇ 2 u ⃗ \mu\nabla^2\vec u μ ∇ 2 u .
Yeh equation kyun matter karti hai? Kyunki yahi har real fluid — blood flow, hawai jahaz ke around air, paani ke pipes, mausam — sab govern karti hai. Aur agar μ = 0 \mu=0 μ = 0 karo toh Euler equation, aur agar u ⃗ = 0 \vec u=0 u = 0 karo toh hydrostatic P = P 0 + ρ g h P=P_0+\rho g h P = P 0 + ρ g h — sab isi ek master equation ke special cases hain. Bas yaad rakho: Mass·accel = Pressure + Viscous + Gravity.