Source near a wall: finding optimal placement

One-line statement: To model fluid flow near an impermeable wall, place a mirror-image source (the "image") symmetrically on the opposite side of the wall, so the combined flow satisfies the no-penetration boundary condition automatically. Key result in words: A source of strength

m

at distance

d

from a flat wall is exactly equivalent (on the fluid side) to that source plus an identical image source at distance

d

on the other side; the wall becomes a streamline. "Optimal placement" refers to choosing

d

— and in richer problems, balancing competing constraints — to achieve a desired flow feature such as a stagnation point, minimal wall velocity, or a target pressure distribution. Where it applies: Two-dimensional ideal (inviscid, incompressible, irrotational) potential flow; groundwater hydrology (wells near impermeable boundaries); electrostatics and heat conduction by mathematical analogy. Who/when: The method of images traces to William Thomson (Lord Kelvin) in electrostatics (1848) and was carried into hydrodynamics by Helmholtz, Kirchhoff, and the classical potential-theory tradition of the late 19th century.

A source in two-dimensional potential flow is a point from which fluid emanates radially at constant volumetric rate per unit depth, called the strength $m$ . When such a source sits near a solid, impermeable wall, the fluid cannot flow through the wall, which imposes a boundary condition: the velocity component normal to the wall must vanish everywhere on it. The method of images solves this by replacing the wall with a fictitious mirror source, reducing a boundary-value problem to free-space superposition. The surprising fact is that an infinite plane wall is reproduced exactly by a single image — no series, no integral — because of the perfect symmetry of the radial source field. "Finding optimal placement" then becomes a clean optimization problem: vary the source's distance from the wall (or its strength, or add a uniform stream) to engineer a specific feature, and the answer often falls out of elementary calculus.

The Core Insight

A flat impermeable wall is a plane of symmetry for the flow. If you reflect every singularity across the wall, the normal velocities induced by each pair cancel on the wall while the tangential velocities add. Hence the wall is a streamline of the doubled, unbounded flow. You never solve a PDE with boundary conditions; you solve free-space superposition and let symmetry enforce the wall.

Formal Statement

Two-Dimensional Source

A line source of strength $m$ (units: $m^{2} / s$ , area per time per unit depth) located at the origin produces a purely radial velocity field. In polar coordinates $(r, θ)$ :

u_{r} = \frac{m}{2 π r}, u_{θ} = 0.

⟳Conservation of volume — flux through any circle of radius

r

\oint u_{r} r d θ = m

, independent of

r

; solve for

u_{r}

Its complex potential (with $z = x + i y$ ) is

w (z) = ϕ + i ψ = \frac{m}{2 π} ln z,

where $ϕ$ is the velocity potential and $ψ$ the stream function. The velocity is recovered from $u - i v = d w / d z = \frac{m}{2 π z} .$

⟳For a holomorphic

w

, the Cauchy–Riemann equations make

ϕ, ψ

harmonic conjugates; differentiate

w = \frac{m}{2 π} ln z = \frac{m}{2 π} (ln r + i θ)

and match real/imaginary parts to

u_{r}, u_{θ}

No-Penetration Boundary Condition

On a solid impermeable wall with unit normal $\hat{n}$ , the fluid cannot cross:

u \cdot \hat{n} = 0 on the wall .

⟳Mass cannot accumulate inside a rigid solid; the normal flux at the surface must be zero (for a stationary wall).

Equivalently, the wall is a streamline: $ψ = const$ along it, since flow runs parallel to streamlines.

Image of a Source in a Plane Wall

Let a source of strength $m$ be located at $(0, d)$ with $d > 0$ , and let the wall be the line $y = 0$ . The flow in the upper half-plane $y > 0$ satisfying $u \cdot \hat{n} = 0$ on $y = 0$ is given by the superposition of the real source at $(0, d)$ and an identical image source of strength $m$ at $(0, - d)$ :

w (z) = \frac{m}{2 π} [ln (z - i d) + ln (z + i d)] .

⟳Place the image by reflection across

y = 0

(i.e.

i d \to - i d

). The combined stream function is symmetric under

y \to - y

, so

\partial ψ / \partial x = v = 0

requires checking; instead verify directly that the normal velocity vanishes on

y = 0

The wall is a streamline, and the upper-half-plane flow is unique.

Why It Holds / Derivation

The Image Cancels Normal Velocity on the Wall

Step 1 — Write the velocity field. Differentiate [4]:

u - i v = \frac{d w}{d z} = \frac{m}{2 π} (\frac{1}{z - i d} + \frac{1}{z + i d}) .

⟳

\frac{d}{d z} ln (z \mp i d) = \frac{1}{z \mp i d}

; sum the two contributions.

Step 2 — Restrict to the wall

y = 0

, so

z = x

(real). Then

(u - i v)_{y = 0} = \frac{m}{2 π} (\frac{1}{x - i d} + \frac{1}{x + i d}) = \frac{m}{2 π} \cdot \frac{2 x}{x ^{2} + d ^{2}} .

⟳Combine over a common denominator;

(x - i d) (x + i d) = x^{2} + d^{2}

, numerator

(x + i d) + (x - i d) = 2 x

is real.

Step 3 — Read off the components. The right-hand side of [6] is purely real, so

u_{y = 0} = \frac{m x}{π ( x ^{2} + d ^{2} )}, v_{y = 0} = 0.

⟳For

u - i v

real, the imaginary part

- v

is zero; equate real parts for

u

Step 4 — Conclude. The wall normal is

\hat{n} = \hat{y}

, and

v = u \cdot \hat{n} = 0

y = 0

. Condition [3] is satisfied identically. By uniqueness of solutions to Laplace's equation with given Neumann data on the boundary (and decay at infinity), [4] is the flow.

■

Source at height $d$ above a wall and its mirror image; the wall (dashed) is a streamline.

William Thomson, Lord Kelvin | 1824–1907

Kelvin introduced the method of images in 1848 to solve electrostatic problems involving a point charge near a grounded conducting plane, replacing the conductor with a mirror charge of opposite sign. The same geometric trick, transplanted to potential flow, gives the source-near-a-wall result — though with an image of the same sign, because the relevant boundary condition (no normal flow) differs from the electrostatic one (constant potential on the conductor). Kelvin's broad program of treating heat, electricity, and fluid motion through the common language of Laplace's equation made such cross-disciplinary transfers natural and routine.

Building Intuition

Imagine the wall as a perfect mirror. Every drop of fluid that the real source pushes toward the wall is met head-on by an identical drop that the image source pushes away from the wall on the other side. Along the mirror line these two impulses are exactly anti-parallel in their normal components, so they cancel — fluid is deflected to slide along the wall rather than through it. The tangential components, by contrast, point the same way and reinforce, which is why the speed along the wall is nonzero and peaks where the geometry is most favorable. The whole construction works because a source's field is perfectly radial and symmetric: reflection maps it onto another genuine source, not some distorted object.

Hermann von Helmholtz | 1821–1894

Helmholtz's vortex theorems and his systematic use of potential theory in fluid dynamics (1858 onward) established the framework in which sources, sinks, vortices, and their images became the elementary building blocks of two-dimensional inviscid flow. The superposition principle that makes image methods possible rests on the linearity of Laplace's equation $\nabla^{2} ϕ = 0$ , the governing equation of irrotational incompressible flow that Helmholtz did so much to popularize among physicists.

Worked Examples

Wall Velocity and Its Maximum

Given: A source of strength

m = 4 π m^{2} / s

at height

d = 2 m

above the wall

y = 0

. Find: (a) the velocity on the wall as a function of

x

; (b) the location and value of the maximum wall speed. Step 1. From [7], the wall velocity is purely tangential:

u (x) = \frac{m x}{π ( x ^{2} + d ^{2} )} = \frac{4 π x}{π ( x ^{2} + 4 )} = \frac{4 x}{x ^{2} + 4} .

Step 2. Maximize:

\frac{d u}{d x} = \frac{4 ( x ^{2} + 4 ) - 4 x ( 2 x )}{( x ^{2} + 4 ) ^{2}} = \frac{4 ( 4 - x ^{2} )}{( x ^{2} + 4 ) ^{2}} = 0 ⟹ x = \pm 2 = \pm d .

Step 3. Evaluate at

x = d = 2

u_{m a x} = \frac{4 ( 2 )}{4 + 4} = \frac{8}{8} = 1 m/s .

Answer: $u (x) = \frac{4 x}{x ^{2} + 4} m/s$ ; maximum speed $u_{m a x} = 1 m/s at x = \pm d = \pm 2 m$ . The fastest flow along the wall occurs directly beneath a 45° line from the source.

Source in a Uniform Stream Near a Wall: Stagnation Point

Given: A uniform stream

U = 3 m/s

flows in the

+ x

direction parallel to the wall

y = 0

. A source of strength

m = 12 π m^{2} / s

sits at

(0, d)

with

d = 1 m

. Use the image to enforce the wall. Setup. The complex potential is the stream plus source plus image:

w (z) = U z + \frac{m}{2 π} [ln (z - i d) + ln (z + i d)] .

Step 1 — Velocity on the wall (

y = 0

z = x

). Add

U

to the tangential result [7]:

u (x) = U + \frac{m x}{π ( x ^{2} + d ^{2} )} = 3 + \frac{12 π x}{π ( x ^{2} + 1 )} = 3 + \frac{12 x}{x ^{2} + 1}, v = 0.

Step 2 — Find stagnation points on the wall (

u = 0

3 + \frac{12 x}{x ^{2} + 1} = 0 ⟹ 3 (x^{2} + 1) + 12 x = 0 ⟹ x^{2} + 4 x + 1 = 0.

Step 3 — Solve the quadratic:

x = \frac{- 4 \pm 16 - 4}{2} = - 2 \pm 3 ⟹ x \approx - 0.27 m or x \approx - 3.73 m .

Answer: Two wall stagnation points at $x = - 2 \pm 3 \approx - 0.27 m, - 3.73 m$ . They lie upstream of the source (negative $x$ ), where the source's leftward-projected outflow opposes the oncoming stream. This is the elementary model of a half-body / Rankine-type fairing sitting on a wall.

Common Misconceptions

The image of a source in a wall is a source, not a sink. This trips up students who remember the electrostatic image of a charge near a grounded conductor, which is an opposite charge. The boundary conditions differ: for a conductor the potential is constant (Dirichlet), forcing opposite sign; for an impermeable wall the normal velocity (a Neumann condition on

ϕ

) is zero, which is satisfied by an image of the same sign and strength. If you mistakenly use a sink, the wall becomes a source/sink line, not a streamline. "Optimal placement" is not a single fixed rule — it depends on the objective. Maximizing wall speed, locating a stagnation point, minimizing peak wall pressure, or keeping a dividing streamline below a given height are different optimization problems with different answers. Always state the objective and constraint before differentiating. A curved or finite wall is generally NOT solved by a single image. Consider a source near a circular boundary instead of a flat one. The correct construction (the Milne–Thomson circle theorem) requires an image at the inverse point

a^{2} / \overset{z_{0}}{ˉ}

plus a compensating source at the center — a single mirror reflection fails because reflecting a radial field across a circle does not produce another single radial source. Treating a circular wall like a flat mirror gives a flow that does not satisfy

u \cdot \hat{n} = 0

on the circle.

Connections

The construction is the fluid-dynamical sibling of the electrostatic Method of Images and shares its mathematical backbone with Laplace's Equation and harmonic functions. Adding a uniform stream connects directly to Rankine Half-Body and Rankine Oval flows. The complex-potential formalism ([2], [4]) is the entry point to Conformal Mapping, where the wall can be straightened or curved by analytic transformations. When the singularity is a vortex rather than a source, the image picks up a sign flip in circulation, linking to Vortex Dynamics and the Milne-Thomson Circle Theorem.

Real-World Applications

Groundwater hydrology. A pumping or injection well near an impermeable bedrock boundary is modeled by an image well of the same strength; near a recharge boundary (constant-head river), the image is a sink, mirroring the electrostatic sign. Used to predict drawdown cones (Theis/Dupuit analysis).
Contaminant plume control. Engineers place an injection source at an optimized distance from an aquifer's no-flow boundary to steer or contain a plume, choosing $d$ to position the dividing streamline.
HVAC and ventilation. Air supply diffusers near a ceiling or floor are idealized as sources near a wall; placement governs the "throw" distance and where the jet attaches (Coandă-like attachment in the inviscid caricature).
Coastal and harbor engineering. Discharge outfalls near a seawall are modeled with image sources to predict near-shore dilution and recirculation.
Microfluidics. Point-injection ports near channel walls in lab-on-a-chip devices use image-source models to design uniform reagent spreading.
Aerodynamic panel methods. Source/sink panels with wall images form the kernel of boundary-element solvers for flow over near-ground bodies (ground effect on wings, race-car floors).
Heat and mass transfer analogs. A heat source near an insulated boundary obeys the same image rule (zero normal heat flux), used in thermal-management layout.

Historical Development

1828 — George Green's Essay develops the potential function and Green's functions, the analytic foundation for image methods.
1848 — William Thomson (Kelvin) introduces the method of images for a point charge near a grounded conducting plane.
1858 — Helmholtz publishes his work on vortex motion, embedding singularities and superposition firmly in fluid dynamics.
1868 — Kirchhoff and contemporaries extend free-streamline and potential-flow theory; complex-variable methods mature.
1880s — W. J. M. Rankine's earlier half-body and oval constructions (1860s–70s) become standard textbook examples of source+stream superposition near and along boundaries.
1906 — Kutta and Joukowski formalize complex-potential aerodynamics, generalizing image and conformal techniques.
1940 — L. M. Milne-Thomson states the circle theorem, giving the systematic image for curved (circular) boundaries.
1960s — The rise of panel methods (Hess & Smith, 1962) turns image-source superposition into a practical numerical tool for arbitrary geometries.

Subtleties & Limits

The clean single-image result is fragile: it holds exactly only for an infinite, flat, perfectly impermeable wall in inviscid, irrotational, incompressible two-dimensional flow. Real subtleties accumulate when these assumptions break:

Finite or cornered walls demand infinite sequences of images (e.g., a source between two parallel walls needs an infinite ladder of reflections, summable via the digamma/cotangent functions), and the series' convergence must be handled carefully.
Viscosity turns the slip wall velocity [7] into a no-slip boundary layer; the inviscid result then describes only the outer flow and feeds boundary-layer matching.
Three dimensions replace the logarithmic potential with a $1/ r$ point source, and "optimal placement" becomes a genuinely 3-D variational problem.
Multi-objective optimality (minimize peak pressure and keep the half-body below a height and fix the discharge) leads to constrained optimization with no closed form — an active design-engineering problem.
A clean open-flavored question: for $N$ sources of fixed total strength near a wall, what placement minimizes the maximum wall shear in the boundary-layer-corrected model? This couples potential flow to viscous theory and resists closed-form solution.