Pressure & Buoyancy · Guide · glidepathscience.org
GuidePressure & Buoyancy
Contents
This Guide explains the ideas and equations behind every Walkthrough step, Concept Challenge question and Predict round, in the order a textbook would take them. Part 1 prepares the Intro track and Part 2 the Advanced track; each takes about fifteen minutes to read. Each section ends with a worked example that is not one of the sim’s own problems.
Part 1 · The essentials
A fluid pushes on everything it touches
A fluid is anything that flows: a liquid such as water, or a gas such as air. A fluid at rest cannot push sideways along a surface; it can only push straight against it. Press your hand into a pool and the water pushes on your palm, the back of your hand and every finger, always at right angles to the skin.
The push per unit area is the pressure. At any point in a fluid at rest it is the same in every direction, or the fluid there would start to move. What changes from point to point is its size, and it changes in one simple way: it grows with depth, because deeper fluid carries the weight of more fluid above it. Almost everything in this sim follows from those two facts, including buoyancy, the difference between the larger push on the bottom of a submerged body and the smaller push on its top. We start with the two quantities every fluid problem uses.
Density, and pressure as force per area
Two blocks of the same size, one of wood and one of lead, have very different masses. What differs is how tightly their mass is packed, their density, the mass per unit volume:
$ρ = m/V$
Its SI unit is the kilogram per cubic meter. Water is about 1,000 kg/m³, the same as 1 g/cm³: a liter of water has a mass of one kilogram. Density belongs to the material, not to the object: cut a block in half and each half keeps the same density. That is why density, not mass, decides whether something floats. Watch the units: grams over cubic centimeters gives a number a thousand times smaller than kilograms over cubic meters.
The second quantity is pressure, a force spread over an area:
$p = F/A$
Its unit, the newton per square meter, is the pascal (Pa), a small unit, so fluid pressures are usually quoted in kilopascals (kPa). Pressure is not force. The same force on a small area is a large pressure, which is why a knife cuts and a snowshoe does not sink, and why a block presses harder on the floor standing on its smallest face.
On its end, the plank’s pressure is also its density times $g$ times its height, $ρgh$ = 750 × 9.81 × 0.40 = 2,943 Pa. A column of anything presses on its base with $ρgh$, whatever its width, and that is the key to the next section.
Pressure grows with depth
Picture a vertical column of water of cross-section $A$, from the surface down to a depth $h$. Its weight, $ρAhg$, pulls it down, and the air pushes down on its top with $p_0A$, where $p_0$ is the atmosphere’s pressure. The water below must push up with just enough to hold up both, $pA = p_0A + ρAhg$. Divide by $A$ and the width disappears:
$p = p_0 + ρgh$
At sea level $p_0$ is about 101.3 kPa, one atmosphere (1 atm). Water adds about 9.8 kPa per meter, so every ten meters adds roughly another atmosphere. A diver feels both the water above and, through it, the air above that: forgetting the atmosphere is the most common slip.
The column argument never used the container’s shape, so the pressure at a given depth is the same in any container, wide or narrow, straight or flared: the hydrostatic paradox. A wide tank holds more water than a thin tube but spreads its weight over a larger bottom. In a funnel that widens upward, the slanted walls hold up the extra water above them. Either way the bottom feels exactly $ρgh$ more than the surface, and connected vessels fill to the same level.
Pressures that include the atmosphere are called absolute. Most gauges read something else.
Gauges, manometers and barometers
A tire gauge is surrounded by air and measures how much harder the tire’s air pushes than the air around it. That difference is the gauge pressure:
$p_g = p - p_0$
A flat tire’s gauge reads zero, but the air inside is still at 101 kPa absolute. Add $p_0$ to a gauge reading before using a law that needs the total pressure.
The simplest gauge is a manometer, a U-tube partly filled with a liquid, one arm on the gas being tested and the other open to the air. A gas that pushes harder than the air drives the liquid up the open side until the extra column, of height $Δh$, balances the difference:
$p_g = ρgΔh$
Dense mercury (13,600 kg/m³) keeps the instrument short; water or alcohol spreads a small pressure over a tall, easily read column. If the open side stands higher, the gas is above the atmosphere’s pressure.
Now close the top of a tube, fill it and turn it over into a dish of the same liquid. The column falls until its weight per area equals the atmosphere’s push on the dish, leaving nearly empty space at the top. That is Evangelista Torricelli’s barometer (1643), $h = p_0/(ρg)$, about 760 mm of mercury at sea level. The same reasoning limits a pump at the top of a well: it cannot pull, only remove the air above the water, and the atmosphere pushing on the well’s surface can hold up only about ten meters of water. Suction is the atmosphere pushing, which is why deep wells put the pump at the bottom.
A barometer works because a push on a liquid’s surface is passed down through the liquid. The next section makes that precise.
Pascal’s principle and the hydraulic lift
Squeeze a sealed bottle full of water and the water’s pressure rises everywhere inside at once, not just under your fingers. Blaise Pascal stated the rule (written about 1653, published 1663): a change in pressure applied to an enclosed fluid is passed on undiminished to every part of the fluid and to the walls of its container. It follows from the depth law: the pressure at any depth is the pressure at the top plus $ρgh$, so raising the top’s pressure raises every depth’s by the same amount. The change is the same everywhere; the pressures themselves still differ with depth.
A hydraulic lift puts this to work. Two pistons of areas $A_1$ and $A_2$, at the same height, sit on one closed volume of oil. A force $F_1$ on the small piston raises the oil’s pressure by $F_1/A_1$, which pushes up on the large piston with $F_2$:
$F_1/A_1 = F_2/A_2$ $A_1d_1 = A_2d_2$
The force is multiplied by $A_2/A_1$. For round pistons the areas go as the radius squared: using the ratio of the radii is the classic mistake. The second equation is the price. The oil cannot be compressed, so the volume pushed out of the small cylinder, $A_1d_1$, appears under the large piston as $A_2d_2$. Multiply the two equations and $F_1d_1 = F_2d_2$: work in equals work out. A lift multiplies force, never energy.
Next: what pressure does to a body sitting in the fluid.
Where buoyancy comes from: Archimedes’ principle
Hold a cube of side $a$ under water. The pushes on opposite sides are at the same depths, so they cancel. The push up on the bottom face beats the push down on the top, because the bottom is deeper by $a$: the difference in pressure is $ρ_{fl}ga$, and over the face’s area $a^2$ it gives a net upward force $ρ_{fl}ga^3$. But $a^3$ is the cube’s volume, so the force is exactly the weight of the water the cube replaces. This is the buoyant force, $F_B$.
The same holds for any shape. The water that once filled the body’s place was at rest, so the surrounding water pushed up on it with exactly its weight, and it pushes on the body the same way. That is Archimedes’ principle:
$F_B = ρ_{fl}V_{sub}g$
The buoyant force equals the weight of the fluid displaced, $w_{fl}$. The body’s own material does not appear: a lead block and an aluminum block of the same size feel the same buoyant force. Nor does the depth, once the body is fully submerged, since water barely compresses. A body hanging from a spring scale under water is held up by the scale and the water together, so the scale reads its apparent weight:
reading $= W - F_B$
On the Buoyancy Bench the reading falls along a straight line as a cylinder of cross-section $A$ is lowered in, with slope $-ρ_{fl}gA$, so the slope gives the liquid’s density. A force sensor that has not been zeroed adds a calibration offset, the same small force, to every reading: it moves the whole line up or down (its intercept) but never tilts it. Zeroing changes the intercept, never the slope.
Archimedes is said to have tested a crown this way: the weight it loses in water is the weight of water of its own volume, so two weighings give its density.
When the buoyant force can equal the weight, no rope is needed at all.
Floating, sinking and balloons
Let go of a body under water and two forces act: its weight, $ρ_{obj}Vg$, and the buoyant force, $ρ_{fl}Vg$. The volume is the same in both, so only the densities decide. Less dense than the fluid, it rises; denser, it sinks; equal, it hovers, which is neutral buoyancy. A 5 kg log floats and a 5 g nail sinks: mass, size and shape do not decide. For a body made of several things, what counts is its average density. A steel ship floats because steel and the air in its hull together are less dense than water; flood the hull and it sinks.
A floating body settles where the buoyant force on its submerged part equals its weight, $ρ_{fl}V_{sub}g = ρ_{obj}Vg$, so the fraction under the surface is
$V_{sub}/V = ρ_{obj}/ρ_{fl}$
It sinks until it displaces its own weight of fluid: the buoyant force on it equals its weight, not more. A larger body of the same material floats with the same fraction submerged, and in a denser fluid the fraction is smaller, so a ship sailing from a river into the sea rises a little. The ratio holds for a rigid body floating freely in one uniform fluid. It does not hold for a body at an oil–water boundary (it displaces some of each), a block pressed onto a dry floor (nothing pushes up on its bottom) or a diver whose air is squeezed (its volume changes with depth).
Air is a fluid too. A hot-air balloon’s envelope is open at the bottom, so the hot air inside is at the outside pressure but less dense. The outside air pushes up with $ρ_{out}Vg$ and the hot air weighs $ρ_{in}Vg$, so the most the balloon can carry, envelope, basket and people included, is
lift $= (ρ_{out} - ρ_{in})V$
Hot air does not rise by itself any more than a cork does: the denser air around pushes it up. By the ideal-gas law, $ρ = pM/(RT)$ with $T$ in kelvins, heating air from 20 °C to 100 °C lowers its density by about a fifth.
Everything so far has taken each fluid’s density to be fixed. Part 2 drops that assumption.
Part 2 · Going further
The hydrostatic equation
Take a thin horizontal slab of fluid of area $A$ and thickness $dy$, with $y$ up. The pressure is $p$ on its bottom and $p + dp$ on its top, and it weighs $ρgA\,dy$. At rest, $pA - (p + dp)A - ρgA\,dy = 0$, the hydrostatic equation:
$dp/dy = -ρg$
With one constant density it gives back $p = p_0 + ρgh$, but it holds for any density that varies with height. For layers of different liquids, integrate piece by piece: each layer adds its own $ρ_igh_i$, and the slope of $p$ against depth changes at each interface. For a density that grows smoothly with depth, such as a salt-gradient solar pond, the gauge pressure is $g∫_0^hρ\,dh$, which for a linear profile is $g$ times the mean density times $h$.
Liquids are not perfectly incompressible. A liquid’s bulk modulus $K$ is the pressure change per fractional change in volume, so $Δρ/ρ ≈ Δp/K$. Water’s $K$ is about 2.2 GPa: 22 MPa, about 2,200 m of seawater, makes it roughly 1 % denser. With $ρ = ρ_0e^{(p - p_0)/K}$ the hydrostatic equation integrates exactly to $Δp = -K\ln(1 - ρ_0gz/K)$.
Water’s density barely changes. Air’s changes completely.
The atmosphere
Air is a compressible fluid. As an ideal gas its density is $ρ = p/(R_sT)$, with $R_s$ = 287.05 J/(kg·K) for air. In the hydrostatic equation that gives $dp/dz = -pg/(R_sT)$: the pressure falls at a rate proportional to itself. At a single temperature the result is exponential, $p = p_0e^{-z/H}$, with the scale height $H = R_sT/g$, about 8.4 km at 15 °C. Each kilometer up loses the same fraction of the pressure, not the same amount; a straight-line drop, as in water, would run out of air near 8 km.
Real air cools as it rises. The International Standard Atmosphere (ISA), the reference behind altimeters and aircraft performance, starts at 101.325 kPa and 15 °C at sea level and cools by $L$ = 6.5 K per km to 11 km, then holds at −56.5 °C to 20 km. With $T = T_0 - Lz$ the hydrostatic equation integrates to
$p = p_0(1 - Lz/T_0)^{g/(R_sL)}$
with $g/(R_sL)$ = 5.256. Colder air is denser, so the ISA’s pressure falls a little faster than an isothermal atmosphere’s at 15 °C. On a hot day the column is warmer and the pressure aloft falls more slowly.
Pressures that grow with depth also push on walls, with a total force and a point where it acts.
Walls and heights
A vertical wall holding back water of depth $H$ feels a gauge pressure that grows from zero at the surface to $ρgH$ at the bottom (the air pushes on both sides and cancels). A strip of width $w$ and height $dy$, at height $y$, feels $dF = ρg(H - y)w\,dy$, and integrating up the wall,
$F = ½ρgwH^2$
The force is the average pressure, half the bottom’s, times the area. Because the deeper water pushes harder, it does not act at mid-depth: its center of pressure, $∫y\,dF/F$, is one third of the way up, $H/3$, at the centroid of the triangle of pressure. Its moment about the bottom edge, $ρgwH^3/6$, grows as the cube of the depth, which is why dams are thickest at the base.
Heights matter in hydraulics too. Pascal’s principle says a change in pressure reaches every part of a fluid, not that the pressure is the same everywhere. Between two piston faces at different heights the oil has weight:
$p_1 = p_2 + ρg(y_2 - y_1)$
An input piston below the output must also hold up the oil column between them.
Integrating pressure over a whole body gives buoyancy, and shows where Archimedes’ principle can fail.
The pressure integral, and weighing in air
The force a fluid exerts on a body is the sum of its pushes over the wetted surface, each along the inward normal:
$\vec F = -∮p\,\hat n\,dA$
For a body surrounded by fluid at rest, the “ghost” argument of Part 1 evaluates it: replace the body by fluid of its shape, which would sit at rest, so the integral equals the weight of fluid displaced and acts through the displaced fluid’s center of mass, the center of buoyancy B. This holds for any shape, in layered fluids too.
The argument needs fluid on every part of the surface. Press a smooth block onto a tank’s floor so tightly that no water gets beneath it, and nothing pushes up on its bottom. The sides cancel and the top is pressed down by the water and, through it, the atmosphere: the net pressure force is downward, often far larger than the block’s weight. Suction cups work this way. A block resting normally has a film of water under it and feels ordinary buoyancy.
Buoyancy in air is small but not always negligible. A balance compares a sample with reference weights, both buoyed up by the air. With air of density $ρ_a$, a sample of density $ρ$ and weights of density $ρ_w$ (steel, 8,000 kg/m³), a reading $m_r$ corresponds to a true mass
$m = m_r(1 - ρ_a/ρ_w)/(1 - ρ_a/ρ)$
A sample less dense than the weights reads light.
For a balloon, air’s buoyancy is the whole story, and a balloon’s gas changes its volume as it rises.
Gases that expand: balloons and divers
A latex weather balloon is free to expand, so its helium stays at the air’s pressure and temperature, and it swells exactly as the air thins. The mass of air it displaces is $pM_{air}V/(RT)$, and since $pV/(RT)$, the moles of helium, is fixed, that mass is $(M_{air}/M_{He})m_{He}$ at every height. The net lift, the mass it can carry besides its helium (its net upward force over $g$), is constant:
$(ρ_a - ρ_{He})V = m_{He}(M_{air}/M_{He} - 1)$
Each kilogram of helium lifts 6.24 kg besides itself, at any height, until the balloon reaches its burst size, near 30 km for a typical sounding balloon.
A sealed balloon with a rigid envelope has a fixed volume, so its lift falls as the air thins, and it rises only until the air’s density equals its own average density, $m/V$: its ceiling. Above it, it is heavier than the air it displaces and sinks back; below, it rises. The air’s density gradient acts as a spring, and the balloon bobs about its ceiling with a period of minutes, slowed by the air it must carry along. A hot-air balloon’s lift, with a fixed volume of heated air, falls with the air’s density.
A diver with air in a buoyancy vest is the opposite of a floating log. Pushed down, a log displaces more water and is pushed back up. A neutral diver who drifts deeper finds the vest’s air squeezed (Boyle’s law, $pV$ constant at a fixed temperature), so the buoyant force falls and the diver sinks faster: neutral buoyancy with a gas pocket is unstable. A Cartesian diver, a vial with a bubble in a sealed bottle, shows it: squeeze the bottle and the diver sinks; let go and it rises.
Floating has its own stability problem, which matters most for ships.
Floating stability: the metacenter
A floating body’s weight acts at its center of gravity, G, and the buoyant force at its center of buoyancy, B, the centroid of the displaced water. Upright, the two are on one vertical line and cancel. Heel the body a little and the side that goes down displaces more water, so B swings toward it. The buoyant force up through the new B and the weight down through G form a couple, which rights the body or rolls it over depending on which side of G the buoyant force’s line passes.
For small heels that line crosses the centerline at a fixed point, the metacenter M. With M above G the couple rights the body. The distance GM is the metacentric height, and with K the keel,
$GM = KB + BM - KG$ $BM = I/V$
$I$ is the second moment of the waterplane’s area about its centerline and $V$ the displaced volume. For a box of beam $b$ at draft $d$, $BM = b^2/(12d)$ and $KB = d/2$. A wide, shallow hull has a large BM, so G may sit well above B and the hull is still stable. The belief that a body is stable only when G is below B confuses floating with a pendulum.
For larger heels we use the righting arm GZ, the horizontal distance between the lines of action of the weight and the buoyant force; the righting moment is $W \cdot GZ$. Against heel, GZ rises with slope GM, peaks and falls to zero at the angle of vanishing stability, past which the ship capsizes. Water sloshing in a half-full tank acts as if G had risen by $ρ_ti_t/(ρV)$, the free-surface effect.
A stable floating body pushed and let go oscillates, which brings us to motion.
Bobbing and added mass
Push a floating block down by $Δy$ and the extra buoyant force, $ρgAΔy$ for a waterline area $A$, pushes it back: a spring of stiffness $k = ρgA$. Its mass equals what it displaces, $m = ρAd$ with $d$ its draft, so it bobs with
$T = 2π\sqrt{d/g}$
The area cancels: a block twice as wide is twice as heavy and twice as stiff. Only the draft matters. Real floats bob more slowly and soon stop, because they must move water as they move and their waves carry energy away.
That first effect is added mass. A body that accelerates through a fluid must also accelerate some of the fluid around it, so it acts heavier. For a sphere in an ideal fluid the extra is half the mass of fluid it displaces, $½ρ_fV$, so a sphere of density $ρ$ let go at rest in a fluid of density $ρ_f$ starts with
$a = (ρ_f - ρ)g/(ρ + C_mρ_f)$, $C_m = ½$
For a heavy body the correction is small; for a light one it dominates. Nearly all of a bubble’s inertia is the water it pushes aside, so it starts at about $2g$, not the enormous value Archimedes’ force over its own mass would give.
The Reference collects the usual mistakes and every equation.
Reference
Common mistakes
- Pressure taken as force. Pressure is force per area: the same weight on a smaller face is a larger pressure.
- The atmosphere forgotten. $p = p_0 + ρgh$ is the absolute pressure; $ρgh$ alone is the gauge pressure.
- Shape thought to matter. At a given depth in a liquid at rest the pressure is the same in any container.
- Suction thought to pull. A pump at the top of a pipe only removes the air; the atmosphere pushes the liquid up, no higher than $p_0/(ρg)$.
- Radii left unsquared. A press multiplies force by the ratio of the areas, $(R/r)^2$, and divides the distance by the same ratio.
- Buoyancy made to depend on the material or the depth. $F_B = ρ_{fl}V_{sub}g$ uses the fluid’s density and the submerged volume only.
- Floating taken to mean $F_B > W$. A floating body at rest displaces exactly its own weight: $F_B = W$.
- The hot air’s own weight forgotten. A hot-air balloon lifts $(ρ_{out} - ρ_{in})V$, not $ρ_{out}V$.
- G below B taken as the condition for stability. A floating body is stable while M is above G, $GM > 0$, even with G well above B.
The equation sheet
| Part 1 | |
|---|---|
| density | $ρ = m/V$ |
| pressure | $p = F/A$ |
| depth | $p = p_0 + ρgh$ |
| gauge | $p_g = p - p_0$; $p_g = ρgΔh$ |
| press | $F_1/A_1 = F_2/A_2$; $A_1d_1 = A_2d_2$ |
| buoyancy | $F_B = ρ_{fl}V_{sub}g$ |
| scale | reading $= W - F_B$ |
| floating | $V_{sub}/V = ρ_{obj}/ρ_{fl}$ |
| hot air | lift $= (ρ_{out} - ρ_{in})V$ |
| Part 2 | |
|---|---|
| hydrostatic | $dp/dy = -ρg$; $Δρ/ρ ≈ Δp/K$ |
| isothermal air | $p = p_0e^{-z/H}$, $H = R_sT/g$ |
| ISA | $p = p_0(1 - Lz/T_0)^{g/(R_sL)}$ |
| wall | $F = ½ρgwH^2$ at $H/3$ |
| heights | $p_1 = p_2 + ρg(y_2 - y_1)$ |
| weighing in air | $m = m_r(1 - ρ_a/ρ_w)/(1 - ρ_a/ρ)$ |
| stability | $GM = KB + BM - KG$, $BM = I/V$ |
| bobbing | $T = 2π\sqrt{d/g}$ |
| added mass | $a = (ρ_f - ρ)g/(ρ + C_mρ_f)$, $C_m = ½$ |
Symbols: $ρ$ density; $p_0$ the atmosphere’s pressure (101.325 kPa at sea level); $h$ depth; $g$ = 9.81 m/s² (9.80665 in the ISA); $V_{sub}$ the submerged volume; $K$ the bulk modulus; $R_s$ air’s gas constant, 287.05 J/(kg·K); $L$ the lapse rate, 6.5 K/km; $w$ and $H$ a wall’s width and depth; $d$ a draft; $C_m$ the added-mass coefficient.
Sources and further reading
- OpenStax, University Physics Volume 1 (Ling, Sanny and Moebs), chapter 14, “Fluid Mechanics,” §14.1–14.4, CC BY-NC-SA 4.0: the Intro track and Part 1 follow its order and notation, in this lab’s own words. Fluids in motion (§14.5–14.7) belong to Fluid Flow & Bernoulli.
- The U.S. Standard Atmosphere 1976 (the ISA); J. N. Newman, Marine Hydrodynamics (MIT Press, 1977), for added mass; A. Biran, Ship Hydrostatics and Stability, for the metacenter and the GZ curve.
- E. Torricelli (1643) and B. Pascal, Traité de l’équilibre des liqueurs (1663). The Field Manual (H) describes the model and what it leaves out.
Pressure & Buoyancy v1.1 · © 2026 Dr. David Meier · CC BY-NC-SA 4.0 · glidepathscience.org