Laplace's equation (elliptic) — physical meaning (steady-state)

WHAT is Laplace's equation?

It is the prototype elliptic PDE. "Elliptic" comes from classifying $Au_{xx}+Bu_{xy}+Cu_{yy}+\dots=0$ by the discriminant $B^2-4AC$ . For Laplace, $A=C=1, B=0$ , so $B^2-4AC=-4<0$ → elliptic (like the equation of an ellipse $x^2+y^2=1$ ).

WHY does "steady-state" produce $\nabla^2 u = 0$ ?

The cleanest derivation starts from the heat (diffusion) equation and asks "what happens when time stops mattering?"

Formula Derivation from the heat equation — from scratch

Step 1. Heat flows from hot to cold, proportional to the temperature gradient (Fourier's law): $\mathbf{q} = -k\,\nabla u$ Why this step? Heat flux $\mathbf{q}$ points down-temperature; $-\nabla u$ points toward colder regions; $k>0$ is conductivity.

Step 2. Conservation of energy: the rate of change of stored heat in a tiny volume equals minus the net heat flowing out: $\rho c\,\frac{\partial u}{\partial t} = -\nabla\cdot \mathbf{q}$ Why this step? $\nabla\cdot\mathbf{q}$ is the net outflow per unit volume (divergence theorem in miniature). If more leaves than enters, the temperature drops.

Step 3. Substitute Step 1 into Step 2: $\rho c\,\frac{\partial u}{\partial t} = \nabla\cdot(k\nabla u) = k\nabla^2 u$ giving the heat equation $u_t = \alpha \nabla^2 u$ with $\alpha = k/(\rho c)$ . Why this step? This is the full time-dependent law; Laplace is its long-time limit.

Step 4. Steady state means nothing changes with time: $\dfrac{\partial u}{\partial t}=0$ . Then $0 = k\nabla^2 u \;\Rightarrow\; \boxed{\nabla^2 u = 0}$ Why this step? When the system has settled, heat-in = heat-out everywhere; no point is gaining or losing energy.

HOW to picture it: the mean-value property

Why it's true (sketch). Define $\bar u(r)=$ average of $u$ on the circle of radius $r$ . Differentiate and use the divergence theorem: $\frac{d\bar u}{dr} = \frac{1}{2\pi r}\oint \frac{\partial u}{\partial n}\,ds = \frac{1}{2\pi r}\iint_{disk}\nabla^2 u\,dA = 0.$ So $\bar u(r)$ is constant in $r$ ; shrinking $r\to 0$ gives $\bar u = u(\mathbf{x}_0)$ .

Figure — Laplace's equation (elliptic) — physical meaning (steady-state)

Boundary conditions — and WHY they make it well-posed

Elliptic problems need conditions on the whole closed boundary (not initial conditions in time, because there's no time!).

Dirichlet: $u$ specified on boundary (fix the edge temperature).
Neumann: $\partial u/\partial n$ specified (fix the heat flux through the edge; insulated ⇒ $=0$ ).

Worked examples

Common mistakes (Steel-manned)

Recall Feynman: explain it to a 12-year-old

Imagine a metal plate. You hold the edges at fixed temperatures — one edge hot, one cold — and wait a long, long time. Eventually the heat stops sloshing around; every spot settles to a temperature that's just the average of the temperatures around it. That settled picture is what Laplace's equation describes. Because each point is the average of its neighbours, you can never have a single super-hot dot in the middle surrounded by cool — it would instantly start sharing its heat and stop being "settled." So the hottest and coldest places are always on the edges you're holding.

Active-recall flashcards

Laplace's equation in symbols

\nabla^2 u = 0

(the Laplacian of

u

is zero)

A function satisfying

\nabla^2 u=0

is called

harmonic

What physical situation does Laplace's equation describe?

A steady-state (time-independent) field, e.g. equilibrium temperature, electrostatic potential, or incompressible irrotational flow

Why does steady-state heat flow give

\nabla^2 u=0

Setting

u_t=0

in the heat equation

u_t=\alpha\nabla^2 u

leaves

\nabla^2 u=0

What is the PDE-classification type of Laplace's equation?

Elliptic (

B^2-4AC<0

with

A=C=1,B=0

)

State the mean-value property

A harmonic function's value at a point equals the average of its values over any surrounding circle/sphere

What does the maximum principle say?

A harmonic function has no interior maxima or minima; extremes occur only on the boundary

What kind of conditions does an elliptic PDE require?

Boundary conditions on the entire closed boundary (Dirichlet or Neumann), not initial conditions

Dirichlet vs Neumann condition

Dirichlet fixes

u

on the boundary; Neumann fixes the normal derivative

\partial u/\partial n

(flux)

u=x^2-y^2

harmonic?

Yes:

u_{xx}+u_{yy}=2-2=0

Solve

u''(x)=0

with

u(0)=10,u(L)=30

u(x)=10+\frac{20}{L}x

(a straight line)

Sign interpretation of

\nabla^2 u>0

at a point

The point is below its neighbours' average; heat flows in, it tends to warm

Connections

Heat equation (parabolic) — diffusion — Laplace is its $t\to\infty$ steady limit.
Poisson's equation — $\nabla^2 u = f$ , Laplace with a source term.
Classification of second-order PDEs (elliptic-parabolic-hyperbolic)
Boundary value problems and well-posedness
Separation of variables for Laplace's equation
Harmonic functions and complex analysis — real/imaginary parts of analytic functions.
Maximum principle
Electrostatics — potential and Gauss's law — $\nabla^2\phi=0$ in charge-free regions.

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, Laplace's equation $\nabla^2 u = 0$ ka matlab hai steady-state — yaani system ab time ke saath change nahi ho raha, sab settle ho chuka hai. Socho ek metal plate jiske edges pe tum fixed temperature laga ke chhod do. Pehle heat idhar-udhar flow karegi, par bahut der baad sab thanda-garam ka jhagda khatam, har point apni jagah par "average of neighbours" ke barabar ho jata hai. Wahi settled picture Laplace describe karta hai.

Iski derivation simple hai: heat equation hai $u_t = \alpha \nabla^2 u$ . Steady-state mein time-derivative $u_t = 0$ , toh bach jata hai $\nabla^2 u = 0$ . Bas! $\nabla^2 u$ batata hai ki ek point apne aas-paas ke average se kitna alag hai. Agar $\nabla^2 u = 0$ , toh point exactly average pe hai, koi net heat flow nahi, isliye steady.

Sabse important property: mean-value property aur maximum principle. Kisi bhi interior point ki value uske around wale circle ka average hoti hai. Isliye andar koi hot spot ya cold spot ho hi nahi sakta — maximum aur minimum hamesha boundary pe milenge. Yeh bilkul logical hai: agar beech mein ek garam dot hota, toh woh apni heat baat raha hota, matlab abhi steady hua hi nahi.

Yaad rakhna: Laplace ek elliptic PDE hai, isliye isme initial condition nahi, boundary condition chahiye — pure closed boundary par (Dirichlet = $u$ fix karo, Neumann = flux fix karo). Time variable hota hi nahi, kyunki yeh toh final settled snapshot hai. Exam mein common galti yeh hoti hai ki log ise time-evolution samajh lete hain — nahi, yeh evolution ka end result hai.

Test yourself — Partial Differential Equations

Connections