Why a Blacksmith Can Fit an Iron Ring 12mm Too Small, Just by Heating It

We all have an intuitive sense of hot and cold -- a kettle of boiling water is obviously hotter than a box of ice -- but physics needs this made precise. Temperature is a relative measure of hotness or coldness: an object with a higher temperature than another is said to be hotter, exactly as 'tall' and 'short' are relative terms. Touch alone can sense temperature roughly, but it is unreliable and far too limited in range for real scientific use. Heat, meanwhile, is the form of energy transferred between two systems, or between a system and its surroundings, purely because of a temperature difference between them -- a glass of ice-cold water left on a table warms up as heat flows in from the room, while a hot cup of tea cools down as heat flows out to the room, until both reach the same temperature as their surroundings. The SI unit of heat energy is the joule (J), while temperature's SI unit is the kelvin (K), with degree Celsius (°C) a commonly used alternative. Two fixed, reliable reference points, the ice point (0°C) and the steam point (100°C) of water under standard pressure, define the Celsius scale, with 100 equal intervals between them; the Fahrenheit scale places these same two points at 32°F and 212°F, 180 equal intervals apart, giving the conversion tF=(9/5)tC+32.

Liquid-in-glass thermometers give slightly different readings depending on the liquid used, since different liquids expand differently. A gas thermometer avoids this: experiments show all gases at low density expand in the same way, obeying the combined ideal-gas equation PV=µRT, where µ is the number of moles and R the universal gas constant (8.31 J/mol/K). Holding a gas's volume constant, its pressure becomes directly proportional to temperature, letting a constant-volume gas thermometer read temperature purely from pressure. Plotting pressure against temperature for several different low-density gases produces straight lines that, remarkably, all extrapolate to the very same temperature at zero pressure: -273.15°C. This universal value is called absolute zero, and it forms the true foundation of the Kelvin scale -- the scale on which -273.15°C is defined as exactly 0 K, related to Celsius simply by T=tC+273.15. Since the Kelvin and Celsius scales share the same size of unit, differing only in where zero sits, this conversion applies at every temperature.

Most substances expand on heating and contract on cooling -- a tightly-screwed metal lid loosens after hot water expands it, mercury rises in a warmed thermometer, and a fully-inflated balloon shrinks in cold water. For a rod of length l, a small temperature change ΔT produces a fractional length change Δl/l=αlΔT, where αl is the coefficient of linear expansion, characteristic of the material. The same idea extends to area (ΔA/A=2αlΔT) and volume (ΔV/V=αVΔT), with a clean relationship between the volume and linear coefficients: αV=3αl. Copper expands about five times more than glass for the same temperature rise, while metals in general have relatively high αl. This is exactly how a blacksmith fits an iron ring onto a wooden cart wheel: if the wheel's rim measures 5.243 m and the ring, at 27°C, measures only 5.231 m across, heating the ring to 218°C expands it by precisely enough to slip over the rim -- where it cools, contracts, and grips tightly, using nothing but the ring's own linear expansion coefficient (αl=1.20x10⁻⁵ K⁻¹ for iron).

Water breaks the usual rule in a genuinely important way: cooling water from room temperature, its volume actually decreases, not increases, until it reaches 4°C -- water has its maximum density at exactly 4°C. Below 4°C, the volume increases again as it approaches freezing, so density decreases. This has a real environmental consequence: as a lake cools toward 4°C, the denser surface water sinks, and warmer, less dense water from below rises to take its place -- but once the surface water cools below 4°C, it becomes less dense again and simply stays at the top, where it eventually freezes. Lakes and ponds therefore freeze from the top down, not the bottom up; if water behaved like ordinary substances instead, lakes would freeze solid from the bottom upward, destroying most of the plant and animal life living in them.

Heating equal masses of different substances by the same temperature rise takes very different amounts of heat -- twice the water needs twice the heat for the same rise, but the same mass of mustard oil needs less heat than water for that same rise. The heat capacity S of a substance is defined as S=ΔQ/ΔT, and dividing by mass gives the specific heat capacity, s=(1/m)(ΔQ/ΔT), the heat needed per unit mass to change temperature by one unit, with SI unit J/kg/K. Water has the highest specific heat capacity of common substances, 4186 J/kg/K -- far higher than iron (450) or copper (386.4) -- which is exactly why water is used as a coolant in car radiators and as a heater in hot water bags. This same property explains real climate patterns: since water warms up (and cools down) more slowly than land, coastal regions experience a moderating sea breeze, while desert regions, lacking large water bodies, see the ground heat up quickly by day and cool quickly by night. Calorimetry, the measurement of heat, rests on a simple principle: when a hotter body is placed in contact with a colder one, with no heat escaping to the surroundings, heat lost by the hot body exactly equals heat gained by the cold body -- exactly the principle used to find an unknown specific heat capacity, such as a 0.047 kg aluminium sphere at 100°C transferring its heat to water and a calorimeter, settling at 23°C, giving the aluminium's specific heat capacity as 0.911 kJ/kg/K.

Heating ice steadily, its temperature rises until it reaches 0°C -- then, remarkably, the temperature stops rising entirely and stays constant until every last bit of ice has melted, even though heat keeps flowing in continuously. This constant-temperature process is melting (or fusion); the reverse is freezing. The same happens at the boiling point: temperature stays fixed at 100°C until all the liquid has become vapour. The heat absorbed or released per unit mass during such a change of state, at constant temperature, is called latent heat, L, given by Q=mL -- Lf for the latent heat of fusion, Lv for vaporisation. For water, Lf=3.33x10⁵ J/kg and Lv=22.6x10⁵ J/kg -- meaning steam at 100°C carries 22.6x10⁵ J/kg MORE heat than water at 100°C, which is exactly why burns from steam are usually far more serious than burns from boiling water at the same temperature. Boiling point itself depends on pressure: it decreases at lower pressure (explaining why cooking is harder at high altitude, where atmospheric pressure is lower) and increases at higher pressure (exactly why a pressure cooker cooks food faster). Converting a full 3 kg of ice at -12°C all the way to steam at 100°C requires heating the ice to 0°C, melting it, heating the resulting water to 100°C, and finally vaporising it -- four separate stages of heat, totalling about 9.1x10⁶ J.

Conduction transfers heat between neighbouring parts of a body through molecular collisions, without any actual flow of matter -- heat one end of a metal rod in a flame, and the other end soon becomes too hot to hold, as heat conducts steadily along its length. For a bar of length L and cross-section A with its ends held at temperatures TC and TD, the steady-state rate of heat flow (the heat current) is H=KA(TC-TD)/L, where K, the thermal conductivity, measures how rapidly a material conducts heat -- silver (406 J/s/m/K) and copper (385) are excellent conductors, while air (0.024), glass wool (0.04), and felt (0.04) are excellent insulators. Joining a steel rod and a copper rod end to end, with the steel end at 300°C and copper end at 0°C, the steady-state junction settles at a temperature found by requiring the same heat current through both rods -- giving 44.4°C for a steel rod of length 15 cm and copper rod of length 10 cm (steel cross-section twice copper's). Cooking pots are often copper-coated on the bottom for exactly this reason, since copper's high conductivity spreads heat evenly for uniform cooking, while builders prefer materials with low conductivity, like glass wool or foam insulation, to keep heat from escaping a room.

Convection transfers heat through the actual bulk motion of a fluid, possible only in liquids and gases -- heated fluid expands, becomes less dense, rises due to buoyancy, and is replaced by colder fluid, which then heats and rises in turn. This natural convection drives the sea breeze directly: land heats faster than water by day, warming and expanding the air above it, which rises and draws in cooler air from the sea; at night the ground cools faster, reversing the cycle. Even the Earth's trade winds arise from natural convection, driven by unequal solar heating between the hot equator and the cooler poles, and modified by the planet's own rotation. Radiation, the third mode, needs no medium at all -- it is how the Sun's energy crosses empty space to reach Earth, and why warmth from a nearby fire is felt almost instantly, faster than conduction or convection could act. All bodies emit thermal radiation; darker bodies absorb and emit it better than lighter ones, exactly why light clothes suit summer and dark clothes suit winter. A perfect radiator of surface area A and absolute temperature T emits at a rate H=AσT⁴ (the Stefan-Boltzmann law), where σ=5.67x10⁻⁸ W/m²/K⁴; real bodies, with emissivity e<1, emit H=AeσT⁴, and a body at temperature T losing heat to surroundings at Ts loses net radiant energy at H=eσA(T⁴-Ts⁴) -- a human body, at skin temperature around 301 K in a 295 K room, radiates roughly 66.4 W this way, over half its total resting energy production. Wien's Displacement Law, λmT=constant (2.9x10⁻³ m K), explains why a heated iron glows dull red, then yellow, then white as it gets hotter, and lets astronomers estimate the surface temperature of stars from their light alone.

Hot water or milk left on a table always cools gradually toward room temperature, never below it. Newton was the first to systematically study this: according to Newton's Law of Cooling, the rate of loss of heat, -dQ/dt, of a body is directly proportional to the temperature difference ΔT=(T2-T1) between the body and its surroundings, valid for small temperature differences. Since dQ=msdT2 for a body of mass m and specific heat s, this gives dT2/(T2-T1)=-Kdt, which integrates to an exponential form: T2=T1+C'e⁻ᴷᵗ. A plot of loge(T2-T1) against time gives a straight line with negative slope, exactly matching this equation, and the rate of cooling is genuinely faster when the temperature difference from the surroundings is larger, slowing down as the body approaches room temperature. This lets the cooling behaviour of a body be predicted precisely: a pan of hot food cooling from 94°C to 86°C in 2 minutes, in a 20°C room, will take only about 42 seconds to cool the much smaller gap from 71°C to 69°C, since the ratio of temperature differences (70°C versus 50°C above room temperature) directly sets the ratio of cooling times.