How this instrument works
Newton's law of cooling states that an object cools, or warms, at a rate proportional to the difference between its own temperature and that of its surroundings. Double the gap and the object sheds heat twice as fast at that instant; halve the gap and the rate halves too. Solving that proportional relationship as a differential equation produces the exponential form T(t) = Tenv + (T0 − Tenv)·e^(−kt): the gap above ambient shrinks by a fixed fraction every equal tick of time, approaching ambient but never quite reaching it.
The cooling constant k folds together everything about the object and its surroundings that a bare temperature reading cannot show: the surface area losing heat, the moving or still air around it, the object's mass and specific heat, even its shape. A thin ceramic mug in a draft carries a higher k than the same coffee poured into an insulated tumbler, so two cups filled from the same pot can read noticeably different temperatures ten minutes later even though both started identical. Newton never measured k directly when he proposed the rule in 1701; he stated only that convective heat loss tracks the temperature difference, leaving k for later users to calibrate from real data.
The law is an approximation, not a universal fact. It assumes the whole object sits at one uniform temperature at every instant, which only holds when heat moves through the material much faster than it leaves the surface — the same lumped-capacitance condition that shows up as a low Biot number. It also assumes convection dominates the heat loss, so a red-hot forging radiating energy, where loss scales with the fourth power of absolute temperature, cools faster early on than this formula predicts. Within those limits, the same curve serves a home cook watching a probe thermometer, a technician timing how long a heat sink needs before a board is safe to handle, and a forensic pathologist reading a body's core temperature to bracket the time since death.
- Enter the Initial temperature — the object's starting reading, in the same degree scale you want the answer in.
- Enter the Ambient temperature — the steady temperature of the surrounding air or fluid the object is cooling toward.
- Enter the Cooling constant k — a rate in units of 1 ⁄ min reflecting the object's surface area, mass, and medium; measure it from your own data or use a documented value.
- Set the Elapsed time in minutes since the object started at its initial temperature.
- Read Temperature at time t — the object's temperature at that moment, found from the exponential decay curve.
Worked example — a 90° cup of coffee after 20 minutes
A 90° cup of coffee sits in a 20° room with a cooling constant of k = 0.05 per minute. After t = 20 minutes, T = 20 + (90 − 20)·e^(−0.05×20) = 20 + 70·e^(−1) = 20 + 70 × 0.367879441 = 20 + 25.7515609 = 45.751560882, or about 45.8°. The cup lost 44.2° in the first twenty minutes but will need far longer than another twenty to lose the remaining 25.8°, because the rate of loss keeps shrinking along with the gap.
The same curve, with a constant calibrated to human physiology rather than a coffee cup, is what a forensic pathologist uses in reverse: measure a body's core temperature once, know the room's ambient temperature, and solve the formula for t instead of T to bracket how long cooling has been underway. A body starting near 37° in a 15° room that has fallen to 25° points to an elapsed time read off the same exponential curve, not a straight-line guess — which is exactly why treating cooling as linear gives a systematically wrong estimate for anything beyond the first hour.
Questions
Why does an object cool quickly at first and then barely at all?
Because the cooling rate is proportional to the temperature gap, not fixed. Right after the object starts hot, the gap to ambient is large and heat leaves fast; as the object nears ambient, the gap shrinks and so does the rate, in the same proportion. That is why a 90° coffee can drop over 40° in the first twenty minutes but take another twenty-plus to lose the last ten — the term e^(−kt) never truly reaches zero, so the object approaches ambient without mathematically equaling it.
What exactly does the cooling constant k represent?
It is a rate constant, in units of inverse time (1 ⁄ min here), that packages everything about the object and its environment a plain temperature reading cannot show — surface area, mass, specific heat, and how briskly air or fluid moves past the surface. A higher k means faster cooling: a mug in a draft might carry k = 0.08/min, while the same coffee in a closed thermos could sit near k = 0.01/min. There is no universal value; it must be measured or looked up for the specific object and setting.
Does Newton's law of cooling also work for warming up?
Yes — the same formula runs in reverse without changing shape. If the initial temperature is below ambient, T0 − Tenv is negative, and the exponential term adds heat back rather than removing it, so the object warms toward Tenv along the identical curve. A cold drink left on a counter follows the same arithmetic as a hot drink cooling, just approaching ambient from the other direction.
When does this formula stop being accurate?
Mainly in two situations: when the object cannot be treated as one uniform temperature throughout, which holds only when heat moves through the material much faster than it escapes the surface, and when radiation rather than convection dominates the heat loss, since radiated power scales with the fourth power of absolute temperature rather than the simple difference this law assumes. A thick roast or a glowing forging cools faster early on than the formula predicts; a thin mug of liquid in still air stays close to ideal.
Can this really estimate time since death?
In outline, yes. Forensic pathologists use a version of this same exponential relationship, with a cooling constant calibrated from body mass, clothing, and environmental conditions, to work backward from a measured core temperature and a known ambient temperature to an elapsed time. Real casework leans on refined models — Henssge's nomogram is the standard one — rather than a generic k, but the underlying physics, proportional heat loss toward ambient, is exactly Newton's law.
Why do the units of the cooling constant matter so much?
Because k carries units of inverse time, and it must match the elapsed time field or the exponent k·t becomes meaningless. A constant of 0.05 per minute needs t entered in minutes; paired with t in hours it would predict wildly overstated cooling. If a source quotes k per second or per hour, convert it to per minute, or convert the elapsed time to match, before entering either value.