The Venturi Effect: Why Squeezing a Pipe Speeds the Flow and Drops the Pressure

The two laws that do all the work

The Venturi effect is the collision of two conservation laws applied to the same tube of fluid. The first is conservation of mass, which for an incompressible fluid in steady flow becomes the continuity equation:

• Continuity: A₁v₁ = A₂v₂ = Q, the volumetric flow rate (m³/s). The same volume that enters the wide pipe per second must leave through the narrow throat per second.
• Because A₂ < A₁, the throat velocity must rise: v₂ = v₁(A₁/A₂). For a circular pipe A ∝ D², so v₂/v₁ = (D₁/D₂)². Halving the diameter quadruples the speed.

The second is conservation of energy along a streamline, written by Daniel Bernoulli in 1738 as Bernoulli's equation:

• Bernoulli: p + ½ρv² + ρgh = constant, where p is static pressure (Pa), ½ρv² is the dynamic pressure, ρ is density (kg/m³), g = 9.81 m/s², and h is elevation (m).
• Each term is an energy per unit volume (J/m³ = Pa). The sum is the fluid's mechanical energy budget; with no friction it cannot change from one point to another.

Put them together on a level pipe (h₁ = h₂). Continuity says v₂ > v₁, so the dynamic-pressure term ½ρv² grows at the throat. To keep the sum fixed, the static pressure p must fall. Faster flow, lower pressure — that single trade is the entire effect.

Deriving the throat pressure drop

Start from Bernoulli between the wide inlet (1) and the throat (2) at the same height:

• p₁ + ½ρv₁² = p₂ + ½ρv₂²
• Rearrange for the pressure difference: Δp = p₁ − p₂ = ½ρ(v₂² − v₁²).
• Use continuity v₁ = v₂(A₂/A₁) to eliminate v₁: Δp = ½ρv₂²[1 − (A₂/A₁)²].

This is why the drop is so aggressive: it scales with the square of the throat velocity, which itself scales with (D₁/D₂)². Take water (ρ = 1000 kg/m³) with a throat speed of 8 m/s: ½ρv₂² = 32 kPa of dynamic pressure available to convert into a pressure deficit. Push the throat faster and the static pressure can dive below zero gauge — even below the vapor pressure of water (about 2.3 kPa at 20 °C), at which point the liquid flashes into vapor bubbles. That is cavitation, the practical ceiling on how hard you can drive a Venturi with a liquid.

Run the standard textbook case — the Venturi meter — backward and you get a flow gauge. Measuring Δp with a manometer between the pipe and the throat lets you solve for the discharge:

• Q = C_d · A₂ · √[ 2Δp / (ρ(1 − (A₂/A₁)²)) ]
• where C_d is a discharge coefficient (typically 0.95–0.99 for a well-made Venturi) that corrects for the small real-fluid losses the ideal derivation ignores.

Why does the fluid actually accelerate?

Continuity tells you the throat velocity must be higher, but it does not explain what pushes each parcel of fluid up to that speed. The honest mechanistic answer is a pressure gradient. As fluid approaches the converging cone, the pressure ahead of it is lower than the pressure behind it. That net forward pressure force does work on the parcel — this is simply Newton's second law, F = ma, written for a fluid element: −∂p/∂x = ρ(dv/dt) along a streamline (the inviscid Euler equation).

So the causal chain runs: the converging wall geometry forces a pressure field in which p is high upstream and low at the throat; that gradient accelerates the fluid; the accelerated fluid, by Bernoulli's bookkeeping, sits at low static pressure. It is a common misconception to say 'the fast flow causes the low pressure' as if speed were the cause — pressure and velocity co-evolve, tied by energy conservation. Equally, the low pressure at the throat is not a suction reaching out to grab things; it is simply that the surrounding higher-pressure fluid pushes matter (fuel, air, a ping-pong ball) toward the throat.

Downstream, in the diverging cone, the process reverses: the pipe widens, continuity slows the flow, and the pressure recovers. A gentle diverging angle (roughly 6–8°) is essential — open it too fast and the boundary layer separates, the flow turns turbulent, and much of the recovered pressure is lost to eddies rather than returned.

The variables that control the effect

Three quantities set the size of a Venturi's punch:

• Area ratio β = D₂/D₁ (the 'beta ratio'). Because velocity ∝ 1/A and Δp ∝ v², shrinking the throat is enormously leveraged. A β = 0.5 throat gives a 4× speed-up; β = 0.33 gives 9×. Small β means big Δp but also big permanent losses and cavitation risk.
• Flow speed / Reynolds number, Re = ρvD/μ. The clean Bernoulli picture holds only when inertial forces dominate viscous ones — typically Re > 10⁴, so a well-behaved turbulent core with thin boundary layers. In slow, viscous, low-Re flow (think honey in a capillary) viscous pressure losses swamp the Bernoulli term and the throat does not develop the expected deficit.
• Density and compressibility. Δp ∝ ρ, so a gas Venturi produces a far smaller pressure drop than a liquid one at the same speed. And once the throat velocity approaches the Mach number M ≈ 0.3, the incompressible assumption fails: the gas density itself changes, and a converging nozzle can 'choke' at M = 1, capping the mass flow no matter how hard you push.

A quick worked feel: air (ρ = 1.2 kg/m³) rushing through a throat at 40 m/s carries ½ρv² = 960 Pa of dynamic pressure — under 1% of an atmosphere, yet more than enough to lift fuel through a carburetor jet or hold a sheet of paper against the flow.

Where the effect earns its keep

The Venturi principle is quietly load-bearing across engineering:

• Venturi flow meters. Standardized (ISO 5167) throats in water mains, oil pipelines, and HVAC ducts measure Q from a single differential-pressure reading, with a permanent head loss of only ~10–20% versus 50%+ for a sharp orifice plate.
• Carburetors and aspirators. The engine's intake air speeds through a throat; the ~1–5 kPa deficit there draws fuel out of the float bowl. A laboratory water aspirator uses the same trick to pull a vacuum down to roughly the water's vapor pressure (~2 kPa absolute) from nothing but a running tap.
• Atomizers, spray guns, and Bunsen burners. Fast air across a small tube's mouth lowers the pressure there and entrains liquid or gas — perfume sprayers and the air-inlet of a Bunsen burner both run on Venturi entrainment.
• Racing aerodynamics. Ground-effect cars and Formula 1 floors shape the underbody as a Venturi: air accelerates in the narrow gap between floor and track, pressure drops, and the higher pressure on top presses the car down — hundreds of kilograms of extra grip with no added weight.
• Medical and process devices. Venturi masks deliver a precise, fixed oxygen concentration by entraining a controlled amount of room air; jet pumps and eductors move slurries with no moving parts.

Subtleties, limits, and the myths to avoid

Bernoulli is not free energy. The equation assumes no friction and no heat exchange along the streamline. Real Venturis lose a few percent of head to viscosity, which is exactly what the discharge coefficient C_d (< 1) and the permanent pressure loss account for. Pretend the losses are zero and your metered flow will be a few percent high.

The height term matters for tall systems. The full Bernoulli sum includes ρgh. For a vertical Venturi or a tall standpipe the ρgh difference between inlet and throat can rival the ½ρv² term (ρg over 1 m of water = 9.8 kPa), and dropping it introduces real error.

Compressibility bites early in gases. The incompressible derivation is trustworthy only up to M ≈ 0.3. Beyond that, use the compressible-flow relations; at the throat of a converging-diverging (de Laval) nozzle the flow can reach exactly Mach 1 and 'choke,' after which lowering downstream pressure no longer increases mass flow — the supersonic regime of rockets and wind tunnels.

The pressure does not 'suck.' A recurring student trap is to imagine the low-pressure throat actively pulling fluid inward. There is no pulling in fluid mechanics — only pushing by pressure. The throat is simply a region the surrounding higher pressure drives things toward. And the flow does not speed up because the molecules 'want to get through the gap'; each parcel accelerates because a real, measurable pressure gradient pushes it, as Newton's second law demands.

Wide pipe vs. narrow throat for water at Q = 5 L/s (5×10⁻³ m³/s), ρ = 1000 kg/m³

Quantity	Wide section (D = 10 cm)	Throat (D = 5 cm)
Cross-section A	7.85×10⁻³ m²	1.96×10⁻³ m²
Velocity v = Q/A	0.64 m/s	2.55 m/s
Dynamic pressure ½ρv²	205 Pa	3.25 kPa
Static pressure change	reference	−3.04 kPa
Velocity ratio v₂/v₁	1×	4× (= (D₁/D₂)²)