Bernoulli Flow Speed Calculator: Two-Point, Pitot & Venturi
Apply Bernoulli's principle for steady, incompressible flow. Calculate velocities, pressures, and head terms using two-point analysis, Pitot tubes, or Venturi/orifice meters. Compare up to 3 scenarios.
Daniel Bernoulli published the principle that bears his name in 1738, in Hydrodynamica. The form most textbooks use today, p + ½ρv² + ρgh = constant along a streamline, is largely Euler's, derived a few years later from Newton's laws applied to fluid elements. The physics: in steady, incompressible, inviscid flow, that sum stays constant along a streamline, so pressure drops where speed rises. The trap most students hit is the continuity equation that lives alongside Bernoulli, A₁v₁ = A₂v₂ for incompressible flow. A 4× area reduction means a 4× speed increase, and any two-point Bernoulli calculation needs both equations together. Skip continuity and the Bernoulli output makes no sense.
Identifying Flow Regime First (Laminar, Turbulent, Transitional)
Before you reach for the Bernoulli equation, ask whether the flow is laminar, transitional, or turbulent. Bernoulli is a streamline equation, and its clean form (no friction term) only behaves well when the streamlines stay clean. In a 50 mm PVC schedule 40 line carrying water at 0.05 m/s, Re comes out near 2,500, sitting right in the transitional band. Pressure drop predictions in that band swing by 30 to 50 percent depending on whether eddies have started to form. That's a problem, and it's one engineers solve by either pushing operating velocity up to keep Re above 4,000, or pulling it down so Re stays under about 1,800 with a margin.
For Pitot tube airspeed work the calculation is different. The flow approaching the probe tip is effectively turbulent free-stream, but the conversion of velocity to stagnation pressure happens along a thin streamline at the probe. As long as the freestream Mach number stays under 0.3, incompressible Bernoulli holds and the regime label of the bulk flow doesn't spoil the local analysis. For a Cessna 172 climbing at 90 knots indicated, that condition is comfortably met. For a Citation jet at 350 knots, the Pitot calculation needs a compressibility correction.
I tell students to write the Reynolds number on the same page as the Bernoulli setup. If Re for the segment comes out under 2,300, switch to the Hagen-Poiseuille pressure drop formula or to a viscous head loss term. If Re is above 4,000, use Darcy-Weisbach with a Moody-chart friction factor. The two-point form of Bernoulli becomes most reliable inside short, smooth, fully developed turbulent runs where viscous losses can be lumped into a single h_loss term.
Field check: Water in 304 stainless tube at 1 m/s through a 25 mm bore gives Re near 25,000. Honey at 25°C through the same line at 1 m/s gives Re near 5. Same geometry, same speed, completely different physics. The fluid wins, not the pipe.
Choosing Geometry: Pipe vs. Plate vs. Cylinder vs. Open Channel
Bernoulli's equation works on a streamline, so the geometry around that streamline drives every modelling choice. Internal pipe flow lets you treat velocity as nearly uniform across the cross section once flow is fully developed, which is why Q = A·v is good enough for most plumbing work. The two-point form between a 5 cm upstream section and a 2 cm contraction in a copper line lives right inside this regime.
Flow over a flat plate or an aircraft wing is a different beast. The streamline along the surface goes through a boundary layer that thickens with distance from the leading edge, so the local velocity used in ½ρv² is the velocity just outside that boundary layer, not the freestream. Pitot tubes on a Cessna are positioned to catch freestream cleanly precisely so this distinction stays small.
Cross flow around a cylinder, like a heat exchanger tube or a smokestack in wind, splits into stagnation flow at the front, accelerated flow on the sides, and separated wake behind. Bernoulli holds along the front-half streamlines, where you can predict suction peaks at the 90 degree shoulders. It fails completely in the wake. Open channel flow brings yet another wrinkle: the free surface acts as a streamline at constant pressure, so Bernoulli simplifies to v²/2g + h = constant along the surface, which is the basis of every weir and spillway calculation.
| Device Type | C_d Range | Permanent Loss | Cost / Complexity | Best For |
|---|---|---|---|---|
| Venturi meter | 0.95 to 0.99 | 5 to 15% of Δp | High / Moderate | Permanent installation, low loss |
| Flow nozzle | 0.92 to 0.99 | 15 to 30% of Δp | Medium / Moderate | High velocity, erosive fluids |
| Orifice plate | 0.60 to 0.65 | 40 to 70% of Δp | Low / Simple | Quick installation, budget jobs |
| Pitot tube | ~1.0 | Negligible | Low / Simple | Point velocity, aircraft airspeed |
The discharge coefficient C_d soaks up vena contracta losses, turbulence, and friction. Lower C_d means the actual flow rate sits well below the ideal Bernoulli prediction, and you have to size the pipe and pump around that.
Boundary Conditions That Change the Answer
Two-point Bernoulli only works once you fix what is happening at each point. If point 1 is the open surface of a swimming pool drain reservoir, the velocity there is essentially zero and the pressure equals atmospheric. If point 2 is the orifice exit at the pump suction, the velocity is whatever the continuity equation gives you and the pressure depends on what comes downstream. Get the boundary wrong and the algebra still runs, but the numbers are nonsense.
Free jet exits to atmosphere are the cleanest boundary you get. Pressure at the jet is atmospheric because the jet has nowhere else to push against. That's how Torricelli's v = √(2gh) drops out of Bernoulli for a tank with a hole, and it's why fire hose nozzles and garden hoses obey the same square-root law. Submerged exits, where the jet discharges into a body of liquid, do not get this simplification because the surrounding fluid pushes back.
Pump and turbine inlets need a different boundary treatment. The energy added or removed has to be inserted as h_pump or h_turbine. A typical ASHRAE chilled water system might run 30 m of head across the pump, and forgetting that term will give you a Bernoulli imbalance that looks like a calculation error when it's really just a missing pump.
Free surfaces in open channel flow act as constant-pressure boundaries, which is why we treat the water surface in a spillway calculation as the streamline of choice. In wind tunnel testing, the freestream far upstream of the model is the boundary, and any blockage correction tries to compensate for the model disturbing this freestream.
Gauge vs. Absolute Pressure: Why It Matters in Engineering
Bernoulli does not care which pressure scale you use, as long as you stay consistent across the equation. If point 1 has gauge pressure and point 2 has absolute, the answer is wrong by 101.3 kPa at sea level. So pick one and stick with it. For most pipe-network engineering work, gauge is more useful because that's what your installed pressure transmitters read, and the ½ρv² and ρgh terms are differences anyway, so the atmospheric reference cancels.
Cavitation is the place where absolute pressure becomes mandatory. Net positive suction head (NPSH) calculations need absolute pressure at the pump inlet, because the threshold for cavitation is the vapor pressure of the fluid at temperature, which is itself an absolute number. Water at 20°C has vapor pressure around 2.3 kPa absolute. A reading of 2.3 kPa gauge at the pump suction sounds harmless until you realize that's 103.6 kPa absolute, far above the vapor pressure, and you have lots of margin. The same number in absolute units would mean the pump is already chewing on vapor bubbles.
Pitot-static airspeed indicators on aircraft work in differential pressure, which is a flavor of gauge. The static port reads ambient atmospheric, and the Pitot port reads stagnation. The difference, ½ρv², is what the airspeed needle responds to. So the indicator works at sea level and at 30,000 feet without any switching, even though absolute pressures differ by a factor of three.
Conservation Laws Applied to the Specific Flow
Bernoulli is a statement of energy conservation along a streamline for an inviscid incompressible fluid. It rides alongside two other conservation laws you can't skip: mass conservation (continuity, A₁v₁ = A₂v₂ for incompressible flow) and momentum conservation, which gives you forces on bends, valves, and reducers. Together they form the working trio for steady flow problems.
Pitot Tube
- Formula: v = √(2Δp / ρ)
- Measures: local velocity at probe tip
- Pros: simple, no permanent loss, cheap
- Cons: point measurement, sensitive to alignment
- Common uses: aircraft airspeed, wind tunnels
Venturi Meter
- Formula: Q = C_d A₂ √(2Δp / ρ(1−β⁴))
- Measures: volumetric flow rate through pipe
- Pros: low permanent loss, stable C_d
- Cons: expensive, needs straight pipe runs
- Common uses: water mains, process plants
Both Pitot and Venturi devices use Bernoulli combined with continuity, but they apply the conservation laws differently. The Pitot tube exploits the stagnation point: continuity forces velocity at the probe tip to zero, and Bernoulli converts that kinetic energy into measurable pressure. The Venturi exploits the throat: continuity forces velocity up at the contraction, and Bernoulli says pressure must drop to compensate. Picking the wrong device for a measurement is almost always a sign that the engineer didn't walk through which conservation law was doing the work.
When you double flow rate Q through a fixed pipe, velocity doubles (continuity), velocity head quadruples (kinetic energy term), and pressure drop across any constriction quadruples (energy split). Friction head loss in turbulent pipe flow scales with v² as well, so doubling throughput in a crude oil pipeline can quadruple pumping power. That quadratic scaling is why Trans-Alaska Pipeline operators run at modest velocities even though the line could pass more crude. Energy cost wins.
When density changes matter, energy conservation expands into the steady flow energy equation with enthalpy and elevation terms. That's the regime where Frank White's Fluid Mechanics takes over from the textbook two-point form, and it's where compressibility corrections enter. For sailing aerodynamics, Frank Schmitt's Aerodynamics of Sails illustrates how Bernoulli plus circulation theory still drives the lift estimate even when the flow over a Genoa headsail is highly three-dimensional.
Common pitfall: writing Bernoulli without first writing continuity. Velocity at point 2 is not a free parameter. It is fixed the moment you choose A₂.
The basic Bernoulli form ignores viscous dissipation, so it only conserves mechanical energy in the idealised case. Real systems lose energy to wall shear, separation, and turbulence. The extended form adds h_loss terms (Darcy-Weisbach for pipe friction, K-factors for fittings) so the conservation accounting balances. ASME MFC-3M and ISO 5167 spell out how to handle these losses for orifice, nozzle, and Venturi installations, including required upstream straight-pipe lengths of around 20 diameters.
Worked Example: Water Through a 5 cm to 2 cm Contraction
Water at 20°C flows through a copper line that contracts from 5 cm bore upstream to 2 cm at the throat. Upstream velocity is 2 m/s, upstream gauge pressure is 200 kPa, the line is horizontal, and density ρ_water = 998 kg/m³. We want the throat velocity and the pressure at the throat.
Step 1: Cross-section areas
A₁ = π(0.025)² = 1.963 × 10⁻³ m². A₂ = π(0.010)² = 3.142 × 10⁻⁴ m². Area ratio A₁/A₂ = 6.25.
Step 2: Continuity for v₂
v₂ = v₁ × (A₁/A₂) = 2 × 6.25 = 12.5 m/s.
Step 3: Bernoulli for Δp
p₁ + ½ρv₁² = p₂ + ½ρv₂². So p₂ = p₁ + ½ρ(v₁² − v₂²) = 200,000 + ½(998)(4 − 156.25) = 200,000 − 75,973 = 124,027 Pa, or about 124 kPa gauge.
Step 4: Reynolds check at the throat
Re = ρv₂D₂/μ = (998)(12.5)(0.020)/(1.002 × 10⁻³) ≈ 249,000. Solidly turbulent, Bernoulli plus a friction term is the right model.
Step 5: Cavitation sanity check
Throat absolute pressure ≈ 124 + 101 = 225 kPa absolute. Vapor pressure of water at 20°C is 2.3 kPa absolute. Plenty of margin, no cavitation risk.
The 76 kPa pressure drop is the price of accelerating the water from 2 m/s to 12.5 m/s. If you connected a U-tube manometer across this section, that's exactly what it would read, give or take a percent for friction. This is the principle a Venturi flow meter exploits: measure Δp, back out Q. It's also the principle behind a Pitot tube on a Cessna 172 sea level at 50 mph difference between static and stagnation pressure: Δp = ½(1.225)(22.4)² ≈ 308 Pa, the dynamic pressure that drives the airspeed needle.
References
The Bernoulli equation as written here assumes steady, incompressible, inviscid flow along a single streamline. Each of those assumptions deserves its own caveat. Real pipe runs have wall friction, so use the extended Bernoulli with h_loss = f(L/D)(v²/2g) and pull f from a Moody chart. Gas flow above Mach 0.3 needs compressibility corrections, often via isentropic flow tables or shock relations. Unsteady flow, like valve closure or pump startup, needs the unsteady momentum equation rather than steady Bernoulli, because water hammer pressures can reach several times the steady value. And separated flow behind bluff bodies or in sudden expansions doesn't have clean streamlines for Bernoulli to live on.
- White, F. M. (2016). Fluid Mechanics (8th ed.). McGraw-Hill. Chapters 3 and 6 cover Bernoulli derivation, limitations, and pipe flow with friction.
- Munson, Young, & Okiishi (2021). Fundamentals of Fluid Mechanics (8th ed.). Wiley. Detailed Bernoulli applications and flow measurement.
- Schmitt, F. K. Aerodynamics of Sails. Practical example of Bernoulli plus circulation theory applied to sail rigs.
- ASME MFC-3M-2004: Measurement of Fluid Flow Using Orifice, Nozzle, and Venturi. Industry standard for C_d correlations and installation.
- ASME PTC 19.5-2004. Performance Test Code on Flow Measurement. The ASME-side counterpart to ISO 5167 covering differential-pressure flowmeter calibration, uncertainty analysis, and pipe-tap installation rules used in performance testing of pumps, turbines, and compressors.
- ISO 5167-1:2022: Measurement of fluid flow by means of pressure differential devices. International standard for differential-pressure flow meters.
- NASA Glenn Research Center: grc.nasa.gov. Educational Bernoulli resources for aerodynamics.
Fixing Bernoulli and Continuity Calculation Errors
Real questions from students stuck on pressure-velocity trade-offs, discharge coefficients, and flow meter selection.
What is Bernoulli's principle?
Bernoulli's principle states that for an inviscid, incompressible fluid in steady flow, an increase in flow speed corresponds to a decrease in pressure or a decrease in the fluid's potential energy. The full equation is p + ½ρv² + ρgh = constant along a streamline, where p is static pressure, ρ is fluid density, v is flow speed, g is gravity, and h is height. Each term has units of pressure (Pa = J/m³). The first term, p, is the static pressure. The dynamic pressure ½ρv² is the kinetic energy carried by each cubic meter of fluid. The hydrostatic term ρgh is the potential energy per unit volume due to height. All three sum to a constant along any streamline as long as the assumptions hold. In a horizontal pipe that narrows from 10 cm to 5 cm diameter, continuity (A₁v₁ = A₂v₂) forces v₂ = 4v₁. If water at 1000 kg/m³ flows at v₁ = 1 m/s in the wide section, then v₂ = 4 m/s. The static pressure difference is Δp = ½ρ(v₁² − v₂²) = ½ · 1000 · (1 − 16) = −7500 Pa. Pressure in the narrow section is 7500 Pa lower. Bernoulli's principle explains how a pitot tube measures airspeed and how a carburetor pulls fuel into an air stream. With caveats, it also describes how an airfoil generates lift. The assumptions break down for viscous flows, for compressible gases above about Mach 0.3, and for any case where the streamline isn't well-defined.
My pressure drop across a Venturi came out negative—doesn't pressure always drop at the throat?
Yes, pressure should drop at the throat (faster flow = lower pressure). A negative Δp in your calculation likely means you swapped p₁ and p₂ or the flow direction. By convention, Δp = p₁ − p₂ where point 1 is upstream (wider) and point 2 is downstream (narrower throat). If you get Δp < 0, check that you're subtracting in the right order. The throat pressure is always lower than upstream for normal flow.
I calculated velocity at the narrow section as the same as the wide section—why doesn't the pipe constriction change anything?
When velocities don't change across a constriction, the cause is almost always missing continuity. Before applying Bernoulli, use A₁v₁ = A₂v₂ to relate velocities. If diameter halves, area drops by 4× (A ∝ D²), so velocity must increase by 4× to conserve mass. Common mistake: treating v₁ and v₂ as independent variables when they're actually linked by continuity. Set up continuity first, then plug the velocity relationship into Bernoulli.
I'm using C_d = 0.97 for an orifice plate but my professor says it should be around 0.62. Who's right?
Your professor. C_d = 0.97 is for Venturi meters with smooth, gradual tapers. Orifice plates have sharp edges that cause flow contraction (vena contracta) and turbulence, giving C_d ≈ 0.60–0.65. Using the wrong C_d gives flow rate errors of 50% or more. Always match C_d to the device type: ~0.97–0.99 for Venturi, ~0.92–0.99 for flow nozzles, ~0.60–0.65 for sharp-edged orifices.
The Pitot tube in my wind tunnel shows 250 Pa differential pressure. What's the airspeed?
Use v = √(2Δp/ρ). For standard air (ρ ≈ 1.225 kg/m³ at sea level, 15°C): v = √(2 × 250 / 1.225) = √408 = 20.2 m/s. That's about 73 km/h or 45 mph. Remember: Pitot tubes measure dynamic pressure (½ρv²), so you're solving for v from the stagnation-static difference. If you're at altitude or different temperature, adjust ρ accordingly—air density varies significantly.
Why does my flow rate calculation give a different answer when I use gauge pressure versus absolute pressure?
For Bernoulli applications, you can use either gauge or absolute pressure—as long as you're consistent at both points. Bernoulli involves pressure differences, so the atmospheric component cancels out. If p₁ and p₂ are both gauge pressures, Δp = p₁ − p₂ is the same as if both were absolute. The problem arises when you mix gauge and absolute at different points. Pick one convention and stick with it throughout.
My calculated velocity is 150 m/s for water in a pipe—that seems way too fast. What went wrong?
Water at 150 m/s would be extraordinary (that's about Mach 0.1 and creates massive friction losses). Check your units: pressure in Pa (not kPa), density in kg/m³ (water = 1000, not 1), and make sure you didn't flip the formula. Also verify your pressure drop is realistic—10 kPa across a Venturi gives ~4.5 m/s for water, which is reasonable. If velocity seems absurd, it usually means a unit error or unrealistic input pressures.
I need to measure airspeed on a drone. Should I use a Pitot tube or a Venturi meter?
Pitot tube, definitely. Venturi meters measure volume flow rate through a closed pipe—they're not designed for open-air velocity measurement. Pitot tubes measure local air velocity at the probe tip, which is exactly what you need for airspeed. They're also lightweight and have no moving parts. Just make sure the probe is aligned with the airflow direction and positioned away from propeller wash or body interference.
The textbook says Bernoulli doesn't apply to turbulent flow, but my pipe has Re = 50,000. Can I still use it?
You can use Bernoulli for turbulent flow, but add a head loss term. The basic Bernoulli equation assumes inviscid flow, which turbulent flow clearly isn't. The fix: use the extended Bernoulli equation H₁ = H₂ + h_loss, where h_loss = f(L/D)(v²/2g) and f comes from the Moody diagram for your Reynolds number and pipe roughness. For short runs with low velocity, losses might be negligible; for long pipelines, they dominate.
I doubled the pipe diameter but velocity only dropped by half, not by four. Why?
That's exactly right. Doubling diameter quadruples area (A ∝ D²), so if flow rate Q stays constant, velocity drops by 4× (v = Q/A). If your velocity only halved, check whether you actually kept flow rate constant. Sometimes people confuse 'doubling diameter' with 'doubling area'—those aren't the same. Doubling D gives 4× the area. If you meant to double the area, then halving velocity is correct.
How do I account for the fire hydrant being 2 meters higher than the water main it connects to?
Include the elevation head (z) terms. Bernoulli is: p₁/ρg + v₁²/2g + z₁ = p₂/ρg + v₂²/2g + z₂. If the hydrant is 2 m higher, set z₂ = z₁ + 2 m (or just z₁ = 0, z₂ = 2). The elevation difference costs about 2 m of head, which equals 19.6 kPa of pressure (ρgh = 1000 × 9.81 × 2). This means 19.6 kPa less pressure is available to accelerate the water, reducing exit velocity compared to a ground-level outlet.