What Makes a Fluid a Fluid
Start here. Apply the same small shear to a solid and to water, and watch only one of them stop deforming.
Skip to the animationA fluid is a substance that cannot resist a shear stress while at rest — apply any shear at all, however small, and it deforms continuously — and Pascal's law, viscosity, the no-slip condition and the boundary layer are all direct consequences of that one inability.
Solid and fluid, precisely
Push a solid sideways and it distorts until its internal elastic stress balances the applied force, then stops. The strain is finite and proportional to the stress — Hooke's law, and the whole of the strength of materials subject.
Apply the same shear to water and it never stops. It deforms continuously, for as long as the force is applied, and there is no threshold below which it behaves like a solid.
This is a much sharper definition than "takes the shape of its container". A block of butter takes the shape of its container too, given enough time, and the difference is exactly whether the deformation ever stops.
| Solid | Fluid | |
|---|---|---|
| Under shear | Deforms finitely, then stops | Deforms continuously |
| Stress relates to | Strain (how much) | Strain rate (how fast) |
| Governing law | τ = G·γ | τ = μ·(du/dy) |
| At rest | Can sustain shear | Cannot sustain any shear |
Consequence one: pressure in a static fluid
- 1Suppose a shear stress existed somewhere in a fluid at rest.
- 2By the definition, that region would be deforming continuously — that is, moving.
- 3But the fluid was stated to be at rest. Contradiction.
- 4Therefore the only stress in a static fluid is normal to any surface considered.
This is why hydrostatic pressure at a point is the same in every direction — Pascal's law — and why the force on a submerged gate is perpendicular to the gate regardless of its angle. The entire statics module rests on this three-line argument.
Consequence two: viscosity
A *moving* fluid does resist — but it resists the rate of deformation rather than its amount. Drag a plate over a thin fluid layer and the force needed is proportional to how fast you drag it, not how far you have dragged it.
Newton's law of viscosity: τ = μ·(du/dy). The velocity gradient du/dy is the rate of shear strain, and the constant μ is the dynamic viscosity. This one equation is the difference between fluid and solid mechanics written down.
- Dynamic viscosity μ
- Units Pa·s. Water is about 1 mPa·s at 20 °C, air about 18 μPa·s — water is roughly 55 times more viscous.
- Kinematic viscosity ν = μ/ρ
- Units m²/s. Because it divides by density, air's kinematic viscosity is about 15 times water's — the opposite ranking to dynamic viscosity, and it is ν that appears in the Reynolds number.
- Temperature dependence
- Liquid viscosity falls sharply with temperature (molecules escape each other's attraction); gas viscosity *rises* with it (more molecular momentum exchange). The mechanisms are different, which is why the trends oppose.
Consequence three: no slip, and the boundary layer
The no-slip condition: fluid in contact with a solid boundary moves at exactly the boundary's velocity. It is an experimental fact rather than a derivation, and it holds for every ordinary flow.
- 1At the wall the velocity is zero, and far from it the velocity is the free-stream value.
- 2So a velocity gradient must exist in between.
- 3A gradient times viscosity is a shear stress — this is skin friction drag.
- 4The thin region containing that gradient is the boundary layer.
It is why a fan blade collects dust: the air at its surface is not moving at all. And it is why boundary layer behaviour — growth, transition, separation — determines drag, stall and most of the loss in a pipe.
Newtonian and non-Newtonian
| Type | Behaviour | Examples |
|---|---|---|
| Newtonian | μ constant; τ ∝ du/dy | Water, air, most oils |
| Shear-thinning (pseudoplastic) | Thins as it is sheared | Paint, blood, ketchup |
| Shear-thickening (dilatant) | Thickens as it is sheared | Cornflour paste, some armour |
| Bingham plastic | Resists until a yield stress, then flows | Toothpaste, drilling mud |
| Ideal fluid | μ = 0 — a modelling fiction | Used where viscosity is negligible |
The syllabus assumes Newtonian throughout, and for water and air that is excellent. Paint is shear-thinning deliberately, so it spreads under a brush and stops running once on the wall. A Bingham plastic technically fails the definition of a fluid below its yield stress — a useful reminder that the definition is a model, not a law of nature.
The other properties, and what each one decides
- Density ρ
- Mass per volume. For liquids essentially constant, which is what the incompressible assumption means; for gases it depends strongly on pressure and temperature.
- Specific weight w = ρg
- Weight per volume. This is the quantity that actually appears in hydrostatic pressure,
p = w·h. - Bulk modulus K
- Resistance to compression. Water's is about 2.2 GPa — high enough to justify treating it as incompressible, and low enough that water hammer is a real and destructive phenomenon.
- Surface tension σ
- Energy per unit area of a free surface. Dominates at small scales — capillary rise, droplet formation, and why an insect can stand on water.
- Vapour pressure
- The pressure at which the liquid boils at a given temperature. If local pressure falls below it — behind a pump impeller or a propeller — bubbles form and collapse violently. That is cavitation, and it destroys metal.
The numbers you will be asked for
- Newton's law of viscosity
τ = μ · (du/dy)
Stress proportional to rate of strain — the defining fluid relationship.
- Kinematic viscosity
ν = μ / ρ
What appears in the Reynolds number. Air's is larger than water's.
- Hydrostatic pressure
p = ρ·g·h
Follows directly from there being no shear in a static fluid.
- Reynolds number
Re = ρ·V·L / μ = V·L / ν
Inertial forces over viscous forces — how much viscosity still matters.
- Bulk modulus
K = −dp / (dV/V)
About 2.2 GPa for water; the reason water hammer exists.
Advantages and disadvantages
Advantages
- One definition generates the statics, kinematics and boundary layer results.
- The Newtonian assumption is excellent for water and air across the whole engineering range.
- Treating liquids as incompressible removes an entire variable from most problems.
- Dimensionless groups let a model in a tank predict a full-size structure.
Disadvantages
- Viscosity varies strongly with temperature, in opposite directions for liquids and gases.
- Non-Newtonian fluids need entirely different constitutive models.
- The incompressible assumption fails for gases above about Mach 0.3, and for water under transient shock.
- Turbulence has no closed-form solution, so most practical results are empirical correlations.
Watch it work
Check yourself
question 1 / 5
One question at a time. Pick an answer to see why it is right or wrong, then move on — there is no score to keep and nothing is saved.