Skip to content
LumenaryAcademy
Thermofluids 75 min

Fluid statics and Bernoulli

Pressure and hydrostatics, manometers, buoyancy and the stability of floating bodies, continuity and Bernoulli's equation — with Venturi meters, Pitot tubes and the truth about lift.

Code tested · 6AI-draftedAwaiting review by a mechanical engineer

Your guides: Chloe · The Designer, Margaret · The Test Engineer, Gideon · The Checker

MotivationH

Fluids at rest and in motion

A dam in the Andes holds back a lake with a wall thinner at the top than at the bottom. A container ship carrying thousands of tonnes of steel floats, while a steel nut sinks. A car's brakes multiply the push of a foot many times through a few millilitres of fluid. A small tube pointing forward on an aircraft tells the pilot the airspeed. A plain constriction in a water main measures the flow through it with no moving parts.

All of these come from two ideas. The first is hydrostatics: in a fluid at rest, pressure depends only on depth. The second is Bernoulli's equation: in a steady flow without friction, pressure, speed and height trade off against each other, exactly as the energy of the roller coaster in Energy and efficiency traded height for speed. This lesson develops both and puts them to work: on gates and dams, manometers, floating barges, flow meters and airspeed probes. It ends with one of the most widely misunderstood stories in engineering, how a wing makes lift. Friction in flowing fluids, which Bernoulli's equation leaves out, is the subject of the next lesson.

DefinitionsH

Pressure and hydrostatics

A fluid is a substance that deforms continuously under any shear stress, however small: it cannot resist being sheared at rest. So in a fluid at rest there is no shear stress at all, only pressure, a normal compressive stress. Pressure at a point acts equally in all directions (a small free-body diagram of a wedge of fluid proves it, since a wedge with unequal pressures on its faces would accelerate), and it acts perpendicular to any surface the fluid touches.

Pressure is measured relative to two references. Absolute pressure is measured from a perfect vacuum; gauge pressure is measured relative to the local atmosphere, which at sea level averages 101.325 kPa. A tyre gauge reading 220 kPa means an absolute pressure of about 320 kPa. Thermodynamic formulas such as pv=RTpv = RT need absolute pressure; structural calculations usually need gauge pressure, because the atmosphere pushes equally on both sides of a wall.

Theorem— The hydrostatic equation

In a fluid at rest under gravity, with zz measured upwards, dpdz=−ρg\dfrac{dp}{dz} = -\rho g. For a liquid of constant density, the pressure at a depth hh below a free surface at pressure p0p_0 is

p=p0+ρgh.p = p_0 + \rho g h .

Proof

Take a vertical cylinder of fluid of cross-section AA between heights zz and z+dzz + dz. Its weight is ρgA dz\rho g A\,dz. The only vertical forces on it are the pressures on its ends, p(z)Ap(z)A upwards and p(z+dz)Ap(z + dz)A downwards (pressures on its sides are horizontal). Equilibrium gives p(z)A−p(z+dz)A−ρgA dz=0p(z)A - p(z + dz)A - \rho g A\,dz = 0, so dp/dz=−ρgdp/dz = -\rho g. With ρ\rho constant, integrating from the surface down a depth hh gives p=p0+ρghp = p_0 + \rho g h.

∎

The pressure rises by about 9.8 kPa for every metre of fresh water, so about one atmosphere for every 10 m: a diver at 30 m is under four atmospheres in total. In air, ρ\rho is about 800 times smaller, and also falls with height, so the atmosphere's pressure halves roughly every 5.5 km.

Two consequences surprise most people. First, the pressure at a given depth does not depend on the shape of the container, only on the depth: the hydrostatic paradox, described by Simon Stevin in 1586. The base of a narrow pipe 10 m tall feels the same pressure as the base of a lake 10 m deep. Second, a pressure applied to a confined fluid is transmitted undiminished to every part of it (Pascal's principle). A hydraulic press or car brake applies a small force F1F_1 to a small piston of area A1A_1; the same pressure acts on a large piston of area A2A_2, which pushes with F2=F1A2/A1F_2 = F_1 A_2/A_1. Energy is still conserved: the large piston moves A1/A2A_1/A_2 as far.

A hydraulic jack is a fine way to lift a car and a dangerous way to hold it up: a seal can leak or a valve be knocked, and the load comes down. That is why workshops require rated axle stands or a proper vehicle lift before anyone goes underneath, and why hydraulic presses carry guards and two-hand controls. A fluid that multiplies force also multiplies the consequences of a failure.

Manometers

The hydrostatic equation turns a column of liquid into a pressure gauge. A manometer is a tube of liquid whose levels respond to the pressures at its two ends. The rule for reading one is simple: start at one end, add ρgh\rho g h for each step down through a fluid, subtract it for each step up, and set the result equal to the pressure at the other end. Points at the same height in the same continuous fluid at rest are at the same pressure.

gaspgaspatmhliquid, density ρ
A U-tube manometer measuring the pressure of gas in a tank. The gas pushes the liquid down in the left leg; the right leg is open to the atmosphere. Points on the dashed line, in the same liquid at the same height, are at the same pressure, so p_gas = p_atm + ρgh. (The gas's own weight is negligible.)

Example— Reading a manometer

A water manometer on a gas supply shows a difference in levels of 200 mm. The gauge pressure of the gas is ρgh=998×9.81×0.200=1958\rho g h = 998 \times 9.81 \times 0.200 = 1958 Pa, about 2 kPa: typical of a domestic gas supply, and far too small to read accurately on a dial gauge built for tyres. For large pressures a denser liquid shortens the column. Mercury, 13.6 times denser than water, was the classic choice (hence "millimetres of mercury" for blood pressure), but because its vapour is toxic it has largely been replaced by electronic pressure transducers, which are calibrated against deadweight testers and other primary standards.

ComputationC

Forces on submerged surfaces

Pressure on a flat submerged surface grows linearly with depth, so the total force is the pressure at the centroid of the surface times its area, F=ρgycAF = \rho g y_c A (in gauge pressure, with ycy_c the depth of the centroid). That force acts not at the centroid but lower down, at the centre of pressure, because the deeper part pushes harder: for a vertical surface, ycp=yc+Ic/(ycA)y_{cp} = y_c + I_c/(y_c A), where IcI_c is the second moment of area of the surface about a horizontal axis through its centroid (the same II you meet for beams in axial load, torsion and bending). The program checks both formulas by integrating the pressure strip by strip, and then looks at a dam.

python
import numpy as np

rho, g = 998.0, 9.81           # fresh water at 20 C (kg/m^3), gravity (m/s^2)

# 1. A vertical rectangular sluice gate, 3.0 m wide and 2.0 m tall, top edge 1.5 m below the surface.
#    Gauge pressure rho g y acts on every strip; add up the strips numerically.
width, top, height = 3.0, 1.5, 2.0
y = np.linspace(top, top + height, 20001)           # depth of each strip (m)
p = rho * g * y                                     # gauge pressure (Pa)
dF = p * width                                      # force per metre of depth (N/m)
F = np.sum(0.5 * (dF[1:] + dF[:-1]) * np.diff(y))   # trapezoidal rule
M = np.sum(0.5 * ((dF * y)[1:] + (dF * y)[:-1]) * np.diff(y))   # moment about the surface
y_cp = M / F                                        # depth of the centre of pressure

# The textbook formulas: F = rho g y_c A, and y_cp = y_c + I / (y_c A)
y_c, A = top + height / 2, width * height
I = width * height**3 / 12                          # second moment of area about the centroid
print(f"gate: F = {F / 1e3:.1f} kN (formula {rho * g * y_c * A / 1e3:.1f} kN)")
print(f"      centre of pressure {y_cp:.3f} m deep (formula {y_c + I / (y_c * A):.3f} m);"
      f" the centroid is at {y_c:.3f} m")

# 2. A dam holding back water 40 m deep: force and overturning moment per metre of length.
H = 40.0
F_dam = rho * g * H**2 / 2                          # triangle of pressure, from 0 to rho g H
print(f"\ndam, 40 m of water: {F_dam / 1e6:.2f} MN per metre of dam, acting {H / 3:.1f} m above the base;")
print(f"     overturning moment about the toe {F_dam * H / 3 / 1e6:.0f} MN m per metre")
print(f"     pressure at the base {rho * g * H / 1e3:.0f} kPa gauge = {rho * g * H / 101325:.1f} atmospheres")
Runs entirely in your browser, in an isolated worker. Nothing is sent to a server.

The 6 m² gate carries 147 kN, the weight of about 15 tonnes, acting 133 mm below its centroid; its hinges and actuator must be sized for both the force and its line of action. The dam carries 7.83 MN on every metre of its length, and because the force grows with the square of the depth, the bottom third of the wall carries more than half of it. That is why gravity dams are triangular in section, thick at the base and thin at the crest, and why their design is checked not only for overturning but also for sliding and for water pressure seeping under the base, which pushes the dam up.

Worked examplesH

Buoyancy and the stability of floating bodies

Theorem— Archimedes' principle

A body wholly or partly immersed in a fluid at rest experiences an upward force, the buoyant force, equal to the weight of the fluid it displaces, acting through the centroid of the displaced volume, called the centre of buoyancy BB.

Proof

Imagine removing the body and filling its submerged volume with the surrounding fluid. That fluid would be in equilibrium, so the pressures on its surface must exactly support its weight, with a resultant acting through its centre of mass, the centroid of the volume. The pressures on the real body's surface are the same, because they depend only on depth. So the real body feels the same force: the weight of the displaced fluid, acting through the same point.

∎

A floating body sinks until it displaces its own weight. A steel ship floats because its hull encloses a large volume of air below the waterline, so it displaces its own weight of water long before the hull is submerged. A body that cannot displace its own weight even when fully submerged, like a steel nut, sinks.

Floating is not enough: the body must also float the right way up. When a ship heels by a small angle, its weight still acts down through its centre of gravity GG, but the underwater shape changes, so the centre of buoyancy moves sideways to B′B', towards the side that has gone down. The vertical line through B′B' meets the ship's centreline at the metacentre MM. If MM is above GG, weight and buoyancy form a couple that rights the ship; if MM is below GG, the couple heels it further.

GMB′BKGZWFBwater surface
A barge heeled by 12°, computed for a beam of five times its draft. The weight W acts down through the centre of gravity G; the buoyant force F_B acts up through the shifted centre of buoyancy B′. The vertical through B′ crosses the centreline at the metacentre M. Because M is above G, the two forces form a couple with arm GZ that turns the barge back upright. B marks the centre of buoyancy when upright and K the keel.

For small angles the geometry gives a beautifully simple result. The distance from BB to MM is BM=I/VBM = I/V, where II is the second moment of area of the waterplane (the ship's footprint at the waterline) about its fore-and-aft centreline and VV is the displaced volume. Measuring heights from the keel KK, the metacentric height is

GM=KB+BM−KG,GM = KB + BM - KG ,

and the righting arm at a small heel angle φ\varphi is GZ≈GMsin⁡φGZ \approx GM \sin\varphi. A positive GMGM means stable. For a box-shaped barge of beam bb and draft TT, KB=T/2KB = T/2 and BM=b2/(12 T)BM = b^2/(12\,T): wide and shallow is very stable.

The program loads a 40 m by 10 m barge with deck cargo whose centre of gravity is 6.5 m above the keel, and watches its stability as the cargo grows. (For a box shape the righting arm at larger angles follows the exact "wall-sided" formula GZ=sin⁡φ (GM+12BMtan⁡2φ)GZ = \sin\varphi\,(GM + \tfrac12 BM\tan^2\varphi), valid while the deck edge stays out of the water and the bottom stays in it.)

python
import numpy as np

# Stability of a rectangular barge in sea water, loaded with deck cargo.
rho = 1025.0                        # sea water, kg/m^3
L, B, D = 40.0, 10.0, 4.0           # length, beam, depth of hull (m)
m_hull, KG_hull = 300e3, 1.6        # empty barge: mass (kg), centre of gravity above keel (m)
KG_cargo = 6.5                      # centre of gravity of the cargo above the keel (m)

print(" cargo (t)   draft T (m)   KB (m)   BM (m)   KG (m)   GM (m)   GZ at 5 deg (m)")
for m_cargo in [0, 200e3, 400e3, 600e3, 800e3]:
    m = m_hull + m_cargo
    T = m / (rho * L * B)                           # Archimedes: displaced mass = barge mass
    KB = T / 2                                      # centroid of the rectangular underwater section
    BM = (L * B**3 / 12) / (L * B * T)              # I of the waterplane / displaced volume
    KG = (m_hull * KG_hull + m_cargo * KG_cargo) / m
    GM = KB + BM - KG
    phi = np.radians(5)
    GZ = np.sin(phi) * (GM + 0.5 * BM * np.tan(phi)**2)   # wall-sided formula
    flag = "" if GM > 0 else "   <- unstable: it will capsize or loll"
    print(f"{m_cargo / 1e3:8.0f} {T:12.2f} {KB:9.2f} {BM:8.2f} {KG:8.2f} {GM:8.2f} {GZ:12.2f}{flag}")
Runs entirely in your browser, in an isolated worker. Nothing is sent to a server.

As cargo is added, two things go wrong at once: the centre of gravity rises, and the barge sits deeper, which reduces BMBM, because the same waterplane must now support a larger volume. The metacentric height falls from over 10 m to 0.03 m at 600 t, and becomes negative before 800 t. The barge would still float, with nearly 1.3 m of freeboard, and still capsize.

DefinitionsH

Continuity and Bernoulli's equation

Now let the fluid move. A streamline is a curve everywhere tangent to the velocity; in steady flow the streamlines stay fixed and each fluid particle follows one. Two laws govern steady flow along a tube of streamlines, such as a pipe.

Definition— continuity equation

Mass is conserved, so in steady flow the mass flow rate is the same through every cross-section: m˙=ρ1A1V1=ρ2A2V2\dot m = \rho_1 A_1 V_1 = \rho_2 A_2 V_2. For an incompressible fluid this is Q=A1V1=A2V2Q = A_1 V_1 = A_2 V_2, where QQ is the volume flow rate in m³/s: where the pipe narrows, the fluid speeds up.

Theorem— Bernoulli's equation

For steady, incompressible, inviscid flow along a streamline, with no shaft work and no heat transfer,

p+12ρV2+ρgz=constant.p + \tfrac12 \rho V^2 + \rho g z = \text{constant} .

Proof

Apply Newton's second law to a small particle of fluid moving along the streamline, of length dsds and cross-section dAdA. With no viscous stresses, the forces along the streamline are the pressure difference −∂p∂sds dA-\frac{\partial p}{\partial s} ds\,dA and the component of weight −ρg dA ds dzds-\rho g\,dA\,ds\,\frac{dz}{ds}. In steady flow its acceleration along the streamline is V dVdsV\,\frac{dV}{ds}. So ρV dV=−dp−ρg dz\rho V\,dV = -dp - \rho g\,dz, and integrating along the streamline with ρ\rho constant gives p+12ρV2+ρgz=constantp + \frac12\rho V^2 + \rho g z = \text{constant}.

∎

Each term is a pressure. pp is the static pressure, the pressure a gauge moving with the fluid would feel; 12ρV2\frac12\rho V^2 is the dynamic pressure; their sum is the stagnation pressure p0p_0, the pressure where the flow is brought to rest; and ρgz\rho g z is the hydrostatic part. Dividing by ρg\rho g gives the same equation in metres of fluid, as heads: pressure head p/ρgp/\rho g, velocity head V2/2gV^2/2g and elevation zz, which is how pump and pipe engineers usually write it.

Every condition in the theorem matters, and most misuse of Bernoulli's equation comes from forgetting one:

  • steady — a flow that is starting, stopping or pulsing needs an extra term;
  • incompressible — fine for liquids, and for gases at low Mach number (VV below about a third of the speed of sound), as the computation below shows;
  • inviscid — friction converts mechanical energy into internal energy, so in long pipes, through valves and in wakes the "constant" falls;
  • along one streamline — comparing points on different streamlines is allowed only if all streamlines start from the same uniform reservoir (irrotational flow);
  • no pumps, fans or turbines in between — they add or remove energy.
ComputationC

Venturi meters, Pitot tubes and draining tanks

The Venturi meter

A Venturi meter is a smooth contraction in a pipe followed by a gentle expansion. Continuity says the fluid speeds up in the narrow throat; Bernoulli says its pressure must therefore drop. Measure the pressure difference and you know the flow. Combining A1V1=A2V2A_1V_1 = A_2V_2 with p1+12ρV12=p2+12ρV22p_1 + \frac12\rho V_1^2 = p_2 + \frac12\rho V_2^2 gives

Q=CdA22 (p1−p2)ρ (1−(A2/A1)2),Q = C_d A_2 \sqrt{\frac{2\,(p_1 - p_2)}{\rho\,\bigl(1 - (A_2/A_1)^2\bigr)}} ,

where the discharge coefficient CdC_d, measured in calibration, corrects for the small effects of friction and the velocity profile: about 0.98 to 0.995 for a well-made Venturi, and about 0.6 for a sharp-edged orifice plate, where the flow separates and much more energy is lost.

12 (throat)flowhmanometer liquid, ρm
A Venturi meter with a differential manometer. The pipe narrows from section 1 to the throat 2 and then widens gently. The faster flow in the throat has a lower pressure, so the heavier manometer liquid is pushed down on the inlet side; the difference in levels h measures p1 − p2 = (ρm − ρ)gh.

The Pitot-static tube

A Pitot tube is a small tube pointing into the flow. Fluid is brought to rest at its open tip, so it senses the stagnation pressure p0p_0. Small holes in the side of an outer tube, parallel to the flow, sense the static pressure pp. Bernoulli between the free stream and the tip gives V=2(p0−p)/ρV = \sqrt{2(p_0 - p)/\rho}. Aircraft, racing cars, wind tunnels and drones all use this; the program also checks how far the incompressible formula can be trusted for air, against the exact compressible relation.

python
import numpy as np

g = 9.81

# 1. A Venturi meter in a water main: 100 mm pipe, 50 mm throat.
#    A mercury-under-water manometer between inlet and throat reads 120 mm.
rho_w, rho_hg = 998.0, 13546.0                 # kg/m^3 at 20 C
d1, d2, h = 0.100, 0.050, 0.120
A1, A2 = np.pi * d1**2 / 4, np.pi * d2**2 / 4
dp = (rho_hg - rho_w) * g * h                  # manometer: the water columns above cancel
V2 = np.sqrt(2 * dp / (rho_w * (1 - (A2 / A1)**2)))   # Bernoulli + continuity
Q_ideal = A2 * V2
Cd = 0.98                                      # typical discharge coefficient of a Venturi
print(f"Venturi: dp = {dp / 1e3:.2f} kPa, throat velocity {V2:.2f} m/s,"
      f" flow {Q_ideal * 1000:.1f} L/s ideal, {Cd * Q_ideal * 1000:.1f} L/s with Cd = {Cd}")

# 2. Pitot-static tubes in air: speed from (stagnation - static) pressure.
rho_air, gam, p_s = 1.204, 1.4, 101325.0       # air at 20 C and sea-level pressure
a = np.sqrt(gam * p_s / rho_air)               # speed of sound, m/s
print(f"\nspeed of sound {a:.0f} m/s")
print("  true V (m/s)   Mach   p0 - p (Pa)   Bernoulli V (m/s)   error")
for V in [10.0, 30.0, 70.0, 100.0, 150.0, 200.0, 250.0]:
    M = V / a
    p0 = p_s * (1 + (gam - 1) / 2 * M**2) ** (gam / (gam - 1))   # compressible, isentropic
    V_b = np.sqrt(2 * (p0 - p_s) / rho_air)                      # incompressible Bernoulli
    print(f"{V:12.0f} {M:8.2f} {p0 - p_s:12.0f} {V_b:16.1f} {(V_b - V) / V:10.1%}")
Runs entirely in your browser, in an isolated worker. Nothing is sent to a server.

The Venturi reads 14.77 kPa and a flow of 10.8 L/s, with no moving parts and very little permanent pressure loss, which is why Venturis are used in water mains, irrigation and process plants, following the procedures of ISO 5167. The Pitot table shows that incompressible Bernoulli is excellent at the speeds of cars, drones and light aircraft (errors well under 1% below Mach 0.2) and still usable at Mach 0.3, but at airliner speeds the error grows to several per cent. Air-data computers on aircraft therefore use the compressible relation.

Draining a tank

Bernoulli between the free surface of a tank and a small hole a depth hh below gives the jet speed V=2ghV = \sqrt{2gh} (Torricelli's law, from the 1640s), the speed of a body falling through hh. As the tank drains, hh falls and so does the flow, which makes the draining time a differential equation: Atank dh/dt=−Cd a2ghA_\text{tank}\,dh/dt = -C_d\,a\sqrt{2gh}. It has an exact solution, but the program solves it numerically with the fourth-order Runge–Kutta method, as you would for a tank of any shape, and checks the answer.

python
import numpy as np

# A tank draining through a hole in its base (Torricelli + continuity), integrated in time.
g = 9.81
A_tank = 1.2                     # m^2: cross-section of the tank
d_hole, Cd = 0.030, 0.61         # sharp-edged 30 mm hole and its discharge coefficient
a_hole = np.pi * d_hole**2 / 4
h0 = 2.0                         # m: initial depth of water

def dhdt(h):
    """Continuity: A_tank dh/dt = -Cd a sqrt(2 g h); Bernoulli gives the jet speed."""
    return -Cd * a_hole * np.sqrt(2 * g * max(h, 0.0)) / A_tank

# Fourth-order Runge-Kutta with a fixed step, recording when the depth passes each mark.
dt, t, h = 0.5, 0.0, h0
marks = [1.5, 1.0, 0.5, 0.1]
print(" depth (m)   time, numerical (s)   time, exact (s)")
k_exact = Cd * a_hole * np.sqrt(2 * g) / (2 * A_tank)   # exact: sqrt(h) = sqrt(h0) - k t
while marks:
    k1 = dhdt(h)
    k2 = dhdt(h + dt / 2 * k1)
    k3 = dhdt(h + dt / 2 * k2)
    k4 = dhdt(h + dt * k3)
    h_new = h + dt / 6 * (k1 + 2 * k2 + 2 * k3 + k4)
    if h_new <= marks[0]:
        t_cross = t + dt * (h - marks[0]) / (h - h_new)          # linear interpolation in the step
        t_exact = (np.sqrt(h0) - np.sqrt(marks[0])) / k_exact
        print(f"{marks.pop(0):9.1f} {t_cross:16.0f} {t_exact:18.0f}")
    t, h = t + dt, h_new
print(f"\ntime to empty completely: {np.sqrt(h0) / k_exact:.0f} s ({np.sqrt(h0) / k_exact / 60:.1f} min);"
      f" the first half of the depth goes in {(np.sqrt(h0) - np.sqrt(h0 / 2)) / k_exact:.0f} s")
Runs entirely in your browser, in an isolated worker. Nothing is sent to a server.

The tank takes 521 s to lose the first metre and 1256 s more to lose the second, because the jet slows as the head falls. The same reasoning sizes the outlets of water towers, fuel tanks and the emergency drains of process vessels.

Common misconceptionsH

Bernoulli and the lift on a wing

Many books, websites and even pilot manuals explain lift like this: air going over the curved top of a wing has further to travel than air going underneath; the two parcels must meet again at the trailing edge, so the top air goes faster, and by Bernoulli its pressure is lower. Almost every step is wrong.

Misconception

"The air over the top must rejoin the air underneath at the trailing edge." No law of physics requires this, and it does not happen. Wind-tunnel smoke pictures and computations show that air over the top of a lifting wing arrives at the trailing edge well before the air that went underneath. The real speed difference is much larger than "equal transit time" would predict, so the theory gets even the size of the lift badly wrong.

Misconception

"Lift needs a curved upper surface." A flat plate at an angle to the flow makes lift; so do paper aeroplanes, symmetric aerobatic wings, and ordinary wings flying upside down at a suitable angle of attack. Camber helps a wing make lift efficiently, but the essential ingredient is that the wing turns the flow.

Misconception

"It is either Bernoulli or Newton." Both are true at once. A wing at an angle of attack, with a sharp trailing edge, deflects a large mass of air downwards (downwash); by Newton's third law the air pushes the wing up. The force is transmitted through the pressure field: the flow curving around the wing needs lower pressure above and higher pressure below, and in the nearly frictionless flow outside the thin boundary layer the speeds and pressures along each streamline obey Bernoulli's equation. Momentum and pressure are two accounts of the same force.

What, then, decides how fast the air goes over the top? The wing's shape and angle, together with one condition that viscosity enforces: the flow must leave the sharp trailing edge smoothly, rather than whipping around it (the Kutta condition). That fixes a net circulation of air around the wing, and the lift per unit span is L′=ρVΓL' = \rho V \Gamma, where Γ\Gamma is that circulation (the Kutta–Joukowski theorem, 1902–1906). Today engineers compute lift with the panel methods and the full simulations of computational fluid dynamics, and confirm it in wind tunnels.

Two more misconceptions

Misconception

"A wider dam, or a bigger lake behind it, means more pressure on the dam." Pressure depends only on depth. A dam holding back a reservoir 40 m deep feels the same pressure whether the reservoir is a narrow gorge or a lake 50 km long; what grows with the length of the dam is the total force, because there are more metres of wall.

Misconception

"Wherever a fluid moves faster, its pressure is lower." Only along a streamline, in steady flow without friction or added energy. A fan or pump makes the air or water both faster and higher in pressure, because it adds energy; a jet leaving a nozzle into the open air moves fast at atmospheric pressure.

ExplanationH

In professional practice

  • Pressure measurement today is mostly electronic: piezoresistive and capacitive transducers, calibrated against deadweight testers that balance a known mass on a precisely made piston, a direct application of p=F/Ap = F/A. Manometers survive as simple, trustworthy references.
  • Flow metering of water, gas and steam for billing and process control relies on differential-pressure meters (Venturis, nozzles and orifice plates) designed to ISO 5167 or ASME's MFC standards, alongside electromagnetic, ultrasonic and Coriolis meters.
  • Ship and offshore stability is computed by naval architects with hydrostatics software over the full range of heel angles and loading conditions, and approved by classification societies against the Intact Stability Code.
  • Pressure testing. Pressure vessels and pipelines are proof-tested before service, and the codes (such as the ASME Boiler and Pressure Vessel Code) prefer testing with water rather than gas: water is nearly incompressible, so it stores little energy, and a failure under hydrostatic test releases a leak rather than a blast. Such tests are done behind barriers by trained staff, never improvised.
  • Dams are designed and monitored under national dam-safety regulations, with instruments measuring the water pressure under and inside the structure throughout its life.
History & global context

A short history

Archimedes of Syracuse stated the principle of buoyancy in the third century BC, in On Floating Bodies, which also analysed the stability of floating paraboloids, an early theory of the metacentre. Simon Stevin of Bruges described the hydrostatic paradox in 1586. Evangelista Torricelli built the mercury barometer in 1643 and derived the law of efflux from a hole; Blaise Pascal arranged for his brother-in-law Florin Périer to carry a barometer up the Puy de Dôme in 1648, showing that air pressure falls with height. Daniel Bernoulli published Hydrodynamica in 1738, relating pressure and speed, and Leonhard Euler derived the general equations of frictionless flow in the 1750s, from which Bernoulli's equation follows. Henri Pitot measured the flow of the Seine with his tube in 1732, Giovanni Battista Venturi studied flow through contractions in 1797, and Clemens Herschel turned the idea into the Venturi meter in 1887. Martin Kutta in Germany and Nikolai Zhukovsky (Joukowski) in Russia explained lift through circulation in the first years of powered flight.

Exercises

Exercises

Use g=9.81g = 9.81 m/s², and ρ=1000\rho = 1000 kg/m³ for fresh water unless told otherwise.

1. Fluid statics and Bernoulli

Warm-upHuman

What is the gauge pressure 10 m below the surface of a freshwater lake (ρ=1000\rho = 1000 kg/m³, g=9.81g = 9.81 m/s²)? Give it in kPa, to the nearest whole number.

2. Fluid statics and Bernoulli

Warm-upHuman

Three containers stand on a table, all filled with water to the same depth of 30 cm: a wide basin, a tall narrow vase, and a funnel-shaped jug that is wide at the top and narrow at the bottom. At which container's base is the water pressure greatest?

Choose one

3. Fluid statics and Bernoulli

CoreHuman

A U-tube manometer filled with water (ρ=1000\rho = 1000 kg/m³) is connected to a duct carrying air; the other leg is open to the atmosphere. The water stands 120 mm higher in the open leg. What is the gauge pressure of the air in the duct, in pascals, to the nearest whole number?

4. Fluid statics and Bernoulli

CoreHuman

A log of uniform density 600 kg/m³ floats in fresh water (1000 kg/m³). What percentage of its volume is below the surface?

5. Fluid statics and Bernoulli

Warm-upHuman

Water flows at 2.0 m/s in a pipe of 100 mm diameter, which then narrows to 50 mm diameter. What is the speed in the narrow pipe, in m/s?

6. Fluid statics and Bernoulli

CoreHuman

In which of these situations can Bernoulli's equation p+12ρV2+ρgz=constantp + \tfrac12\rho V^2 + \rho g z = \text{constant} reasonably be applied between the two points named? Select all that apply.

Choose all that apply

7. Fluid statics and Bernoulli

CoreComputational

Write venturi_flow(d1, d2, dp, rho=1000.0, Cd=0.98) returning the volume flow rate in m³/s through a Venturi meter with inlet diameter d1 and throat diameter d2 (both in m), a measured pressure difference dp = p1 - p2 (Pa), fluid density rho (kg/m³) and discharge coefficient Cd.

Runs entirely in your browser, in an isolated worker. Nothing is sent to a server.

8. Fluid statics and Bernoulli

StretchComputational

Write barge_stability(L, B, mass, KG, rho=1025.0) for a box-shaped barge of length L and beam B (m) with total mass mass (kg) and centre of gravity KG metres above the keel, floating in water of density rho. Return (T, GM): the draft and the metacentric height, both in metres. (You may assume the barge is deep enough not to sink.)

Runs entirely in your browser, in an isolated worker. Nothing is sent to a server.

Further reading and resources

  • NASA Glenn Research Center: Beginner's Guide to AeronauticsPages on pressure, Bernoulli's equation, Pitot tubes and lift, including a careful discussion of incorrect theories of lift.
  • How do wings work? · Holger Babinsky · 2003Physics Education 38(6). A short, clear article explaining lift through streamline curvature, written for teachers.
  • Understanding Aerodynamics: Arguing from the Real Physics · Doug McLean · 2012Wiley. A book-length treatment by an aerodynamicist, with a thorough account of lift and its misconceptions.
  • ISO, the International Organization for StandardizationPublishes the ISO 5167 series on measuring flow with orifice plates, nozzles and Venturi tubes.
  • International Maritime OrganizationIts Intact Stability Code sets the stability criteria that ships must meet.
  • ASMEPublishes the MFC standards on flow measurement and the Boiler and Pressure Vessel Code, which governs pressure testing.