Bénard–Marangoni Convection: Why Heated Thin Films Draw Hexagons (Field Guide)

A thin liquid layer can look perfectly calm right up until it suddenly starts organizing itself into a tiled city of cells.

Warm spots appear on the surface, fluid races sideways, cooler regions pull harder than hotter ones, and the whole film locks into rolls, hexagons, squares, or time-dependent flicker.

That instability is Bénard–Marangoni convection: convection driven not mainly by buoyancy, but by surface-tension gradients along a free surface.

One-Line Intuition

If one patch of a liquid surface gets slightly warmer, its surface tension usually drops; nearby cooler surface pulls harder, dragging fluid sideways, which drives return flow underneath and can amplify the original temperature variation into an organized convection pattern.

The Core Trick: Surface Tension Is a Force Field, Not Just a Number

People often meet surface tension as a static idea:

• droplets bead up,
• soap changes wetting,
• insects walk on water.

But once surface tension varies along the interface, it stops being just a passive property and becomes a driver of shear flow.

For most liquids,

• hotter surface → lower surface tension,
• cooler surface → higher surface tension.

A higher-surface-tension patch pulls on neighboring liquid more strongly than a lower-surface-tension patch. So the surface flow tends to run:

• away from warmer regions
• toward cooler regions.

That surface motion is the seed of the instability.

How the Feedback Loop Works

Imagine a shallow liquid layer heated from below, with a free surface on top.

Now give the surface a tiny temperature fluctuation: one little patch becomes slightly warmer than its surroundings.

1. That warm patch gets lower surface tension.
2. Nearby cooler surface pulls liquid away from the warm spot.
3. Mass conservation demands a return flow in the bulk.
4. That recirculation can pull warmer liquid upward from below and/or redistribute heat so the original surface contrast strengthens.
5. The tiny perturbation grows into a real convection cell.

So this is a classic pattern-forming instability:

• a tiny fluctuation appears,
• the system does not erase it,
• the system amplifies it.

Why Thin Layers Change the Story

This is the crucial contrast with ordinary Rayleigh–Bénard convection.

Rayleigh–Bénard

Driven mainly by buoyancy:

• hot fluid below is lighter,
• cold fluid above is heavier,
• gravity wants to overturn the layer.

Bénard–Marangoni

Driven mainly by surface-tension gradients:

• the free surface itself creates tangential stress,
• gravity is not the main actor,
• thin layers often make the surface-driven mechanism dominate.

That is why Bénard’s early hexagonal-cell experiments were historically confusing. For a while, people tried to explain them as pure buoyancy convection. Later work — especially Pearson’s 1958 analysis — showed that in many of those thin-layer/free-surface cases, surface tension was the missing engine.

A good memory hook:

• deep layer + strong gravity story → think Rayleigh,
• thin layer + free surface + strong interface physics → think Marangoni.

The Dimensionless Control Knob: Marangoni Number

The instability is usually summarized by the Marangoni number,

Ma ~ (|dσ/dT| · ΔT · d) / (μ · α)

where, roughly:

• |dσ/dT| = how strongly surface tension changes with temperature,
• ΔT = imposed temperature difference,
• d = layer thickness,
• μ = dynamic viscosity,
• α = thermal diffusivity.

Interpretation:

• numerator = how hard surface-tension gradients try to drive flow,
• denominator = how strongly viscosity and thermal diffusion smear it out.

So:

• bigger temperature sensitivity,
• bigger temperature drop,
• thicker thin-film layer,
• lower viscosity,
• lower thermal diffusivity

all make surface-driven convection easier to trigger.

For Pearson’s classic idealized free-surface problem, the primary instability appears at a critical Marangoni number around 79.6 (with boundary-condition caveats). Real experiments shift that threshold because of evaporation, heat loss to air, deformable surfaces, contamination by surfactants, and coupling to the gas above.

So Ma ≈ 80 is a useful landmark, not a universal law.

Why the Patterns Often Look Like Hexagons

Near onset, the system wants a planform that tiles space while satisfying the symmetry and feedback constraints of the flow.

That often gives you:

• hexagonal cells,
• sometimes rolls,
• sometimes squares,
• and at stronger forcing, oscillatory or chaotic states.

Hexagons are famous because they are visually striking and energetically convenient for packing similar convection cells in a plane. But they are not mandatory.

Pattern selection depends on things like:

• layer depth,
• surface deformability,
• evaporation strength,
• fluid properties (especially Prandtl number),
• lateral boundary conditions,
• whether buoyancy and Marangoni effects are cooperating or competing.

That is why “Bénard cells” are a useful visual label, but not a full diagnosis.

Evaporation Makes It More Real — and More Annoying

In lab cartoons, the top surface is just a neat free interface.

In real life, thin films often evaporate. That matters because evaporation can:

• cool the surface,
• sharpen temperature gradients,
• change film thickness in time,
• introduce concentration gradients as mixtures dry,
• add solutal Marangoni effects on top of thermal ones.

This is why drying coatings, inks, polymer films, and volatile mixtures love producing beautiful but inconvenient self-organized patterns.

In other words: Bénard–Marangoni convection is not just a fluid-dynamics party trick. It is one reason a thin film can dry into something you did not intend.

Why Microgravity People Care

Because the mechanism is surface-driven rather than gravity-driven, it becomes especially visible in microgravity.

On Earth, buoyancy can mask or compete with Marangoni flow. In low gravity:

• buoyant overturning weakens,
• surface-tension-driven transport stands out,
• thermocapillary flow becomes easier to isolate.

That makes Bénard–Marangoni physics important in:

• space materials processing,
• crystal growth,
• floating-zone experiments,
• microgravity fluid management.

It is one of those rare cases where removing gravity does not just simplify the experiment — it changes which instability gets to be the boss.

Practical Places It Shows Up

1. Drying paint, coatings, and thin films

Small thermal/compositional gradients can create cellular thickness variation, streaks, or texture defects.

2. Welding, soldering, and laser melt pools

Surface-tension-driven flow can reorganize heat and solute transport, changing pool shape, mixing, and defect formation.

3. Crystal growth and semiconductor processing

Thermocapillary flow can be useful for mixing, but it can also destabilize interfaces and degrade uniformity.

4. Microfluidics

At small scales, surface forces become disproportionately important, so Marangoni stresses can steer flows without moving parts.

5. Soap/surfactant contamination problems

A tiny amount of surface-active material can radically change the surface-tension landscape and therefore the flow pattern.

The Big Operator Lesson

When a shallow liquid layer develops a mysterious cellular pattern, ask this first:

Is the interface actively pulling fluid around?

If yes, do not default to a buoyancy-only explanation.

A quick diagnostic checklist:

• Is there a free surface?
• Is the layer thin?
• Does surface tension likely vary strongly with temperature or concentration?
• Is evaporation involved?
• Do patterns persist even when buoyancy seems too weak to explain them?
• Does changing surfactant contamination or gas-side cooling radically alter the pattern?

If several answers are yes, Bénard–Marangoni physics is probably in the room.

What This Instability Is Not

1. “Just ordinary boiling.”

No. You do not need phase-change bubbles. The key driver is tangential stress from surface-tension gradients.

2. “Just Rayleigh–Bénard with a prettier name.”

No. Buoyancy and Marangoni effects can coexist, but they are distinct mechanisms.

3. “Only a lab curiosity.”

Also no. It matters in coatings, melts, crystal growth, space experiments, and any process where thin free-surface layers carry thermal or concentration gradients.

4. “Any hexagonal pattern on a liquid means Marangoni convection.”

Not necessarily. Pattern shape alone is not proof. Mechanism matters.

Why It’s Such a Good Complex-Systems Example

Bénard–Marangoni convection is a wonderful reminder that:

• a nearly invisible material property gradient can overpower naive intuition,
• interfaces are active dynamical objects,
• thin systems often obey different rulers than bulk systems,
• clean geometric patterns can be the first visible sign of a feedback loop crossing threshold.

It is the kind of phenomenon that makes fluid dynamics feel almost architectural:

a warm, shallow film quietly discovers that the easiest way to move heat is to start drawing tiles.

Tiny Mental Picture To Keep

If Rayleigh–Bénard convection is:

• “hot stuff rises.”

then Bénard–Marangoni convection is:

• “cooler surface pulls harder.”

That one sentence gets you surprisingly far.

References / Pointers

• Pearson, J. R. A. (1958). On convection cells induced by surface tension. Journal of Fluid Mechanics.
• Scriven, L. E., & Sternling, C. V. (1960). The Marangoni effects. Nature.
• Koschmieder, E. L. Bénard Cells and Taylor Vortices. Cambridge University Press.
• University of Toronto, Experimental Nonlinear Physics Group: Rayleigh–Bénard and Bénard–Marangoni convection.
• Reviews on secondary instabilities in surface-tension-driven Bénard–Marangoni convection for hexagon/square/oscillatory pattern selection.