The Gold Standard of Thermodynamics | Finite Bodies, Entropy Generation and the Direction of Heat Flow

Take two bodies with constant heat capacities Cₕ and C꜀. Heat capacity is the energy required for one kelvin of temperature change under the model’s stated constraints. We assume each body has a uniform equilibrium temperature initially and finally, undergoes no phase change or chemical reaction, and can be represented by a constant C over the interval.

The combined outer boundary is insulated. No matter crosses it, and no work crosses it during direct contact. Ignore the energy and entropy changes of the connector and enclosure. These assumptions make the two bodies together an isolated composite system.

The bodies individually are not isolated: they exchange energy with each other. That distinction matters because one body’s entropy may decrease while the combined entropy increases.

Write the boundary before selecting an equation. “Insulated” refers to transfer through the chosen outer boundary. It does not prevent internal heat flow or require each part to retain its original temperature.

Contents · Next chapter

Choose Cₕ = C꜀ = 1,000 joules per kelvin. Let the initial hot temperature be 360 kelvin and the initial cool temperature 300 kelvin. At final thermal equilibrium both have temperature Tᶠ.

The hot body’s energy change is 1,000(Tᶠ − 360). The cool body’s is 1,000(Tᶠ − 300). Their sum is zero because no energy crosses the outer boundary.

Solving gives 2,000Tᶠ = 660,000, hence Tᶠ = 330 kelvin. The hot body loses 30,000 joules and the cool body gains 30,000 joules.

The general constant-capacity result is Tᶠ = (CₕTₕ + C꜀T꜀)/(Cₕ + C꜀). Equal capacities make it an arithmetic mean; unequal capacities require weights.

This is an endpoint calculation. It does not tell us how quickly equilibrium is approached. That timing depends on heat-transfer paths and their changing temperature differences. Energy balance and transfer-rate modelling answer connected but different questions.

Contents · Previous chapter · Next chapter

Entropy S is a thermodynamic state function. Its change between equilibrium states can be calculated along a suitable imagined reversible path, even when the actual path is irreversible.

For our constant-heat-capacity model, a small reversible energy input is C dT. Dividing by absolute temperature and summing along the reversible path gives ΔS = C ln(Tᶠ/Tⁱ). Here ln is the natural logarithm, and temperatures in the ratio must be absolute temperatures in kelvin.

The ratio is dimensionless. Multiplying its logarithm by C gives joules per kelvin, the unit of entropy change. Heat capacity and entropy can share a unit without being the same physical quantity.

The reversible path is a calculation device. We are not claiming that direct contact between bodies at different temperatures is reversible. State-function reasoning lets us calculate each entropy change from the endpoints while keeping the irreversibility of the real process intact.

Contents · Previous chapter · Next chapter

For the hot body, ΔSₕ = 1,000 ln(330/360), approximately −87.01 J/K. The logarithm is negative because the ratio is less than one. Cooling reduces this body’s entropy under the stated model.

For the cool body, ΔS꜀ = 1,000 ln(330/300), approximately +95.31 J/K. The increase is larger in magnitude than the hot body’s decrease.

The total energy change is zero, while the total entropy change is positive. The two balances do not conflict because they describe different quantities.

Since no entropy crosses the insulated outer boundary and the connector and enclosure contributions are neglected, the combined entropy increase is the entropy generated within this model. Entropy generation is not energy generation. The 30,000 joules transferred internally have already been fully accounted for.

Contents · Previous chapter · Next chapter

A common error is to write: “The combined system receives no heat, so ΔS = Q/T = 0.” The formula has been applied outside its conditions. Zero external heat does not imply zero entropy generation inside the system.

For a reversible isothermal transfer, entropy change can be expressed as Qᵣₑᵥ/T. Our actual process has finite temperature differences and changing temperatures. It is not a single reversible isothermal heat transfer.

Even separately dividing 30,000 joules by each body’s initial temperature is not exact for these finite bodies. Those temperatures change as energy moves. The logarithms account for that changing temperature along the reversible calculation path.

For comparison, a deliberately different model transfers 3,000 joules between ideal reservoirs fixed at 360 and 300 kelvin. Its combined entropy increase is 3,000/300 − 3,000/360 = 1.667 J/K. Those reservoirs are assumed not to change temperature. Keep the fixed-reservoir calculation separate from the finite-body model rather than borrowing its convenient denominators.

Contents · Previous chapter · Next chapter

Start with the two equal-capacity bodies both at 330 kelvin. Imagine that, while remaining isolated and doing no work, one spontaneously reaches 360 kelvin and the other 300 kelvin.

The energy balance permits those endpoints: one gains 30,000 joules and the other loses the same amount. But the combined entropy change would be −8.30 J/K, the negative of the previous result.

For an isolated system, that decrease violates the second law. Energy conservation is necessary, but it is not a sufficient test of whether a proposed process can occur.

A refrigerator can create a temperature difference with work input and a suitable wider energy and entropy account. That is a different process with a different boundary. The second law does not forbid cooling one region; it constrains the full process, including the other regions and transfers needed to accomplish it.

The useful habit is to test both balances. If a claim reports only that “all the energy adds up,” ask what establishes the direction and the physical possibility.

Contents · Previous chapter · Next chapter

Keep initial temperatures at 360 and 300 kelvin, but make the hot body’s heat capacity 2,000 J/K while the cool body remains 1,000 J/K.

The final temperature is (2,000×360 + 1,000×300)/3,000 = 340 kelvin. The hot body loses 2,000×20 = 40,000 joules; the cool one gains 1,000×40 = 40,000 joules.

Their entropy changes are 2,000 ln(340/360), approximately −114.32 J/K, and 1,000 ln(340/300), approximately +125.16 J/K. The combined increase is about 10.85 J/K.

The final temperature is closer to the hotter starting temperature because that body has the greater heat capacity. Averaging the two starting temperatures without weights would fail the energy balance.

This changed-context check also prevents a memorised “hot entropy loss equals cold entropy gain” rule. The energies exchanged are equal in magnitude; the entropy changes generally are not. Heat capacity, temperature and the process boundary all remain part of the account.

Contents · Previous chapter · Next chapter

Return to equal capacities of 1,000 J/K. Compare the 360-and-300-kelvin start with a new 345-and-315-kelvin start. Both have the same initial total energy under this model, and both equilibrate directly at 330 kelvin.

For the narrower starting difference, the total entropy increase is 1,000 ln(330/345) + 1,000 ln(330/315), approximately 2.07 J/K. This is smaller than 8.30 J/K for the wider starting difference.

The initial total energy and final temperature match across the cases. Their initial entropy and generated entropy differ. A total energy figure alone therefore does not describe every thermodynamically relevant feature of the starting arrangement.

This comparison has controlled equal capacities and a fixed arithmetic-mean temperature. It should not become an unqualified rule that any pair with a larger temperature difference must generate more entropy under every possible change of mass, material or boundary. State what was held fixed and what changed. That makes the comparison transferable without making it overbroad.

Contents · Previous chapter · Next chapter

Now imagine an ideal reversible heat engine operating between the original finite bodies, rather than simply connecting them thermally. Its net work leaves the combined boundary. The overall process is insulated, but it is no longer isolated because energy crosses as work.

In the ideal reversible limit, total entropy generation is zero. For equal constant capacities, the final common temperature Tʳ satisfies ln(Tʳ/360) + ln(Tʳ/300) = 0. Thus Tʳ is the geometric mean: square root of 360×300, approximately 328.63 kelvin.

The final temperature is lower than the 330 kelvin reached by direct contact because some of the initial energy has left as work. The ideal net work is 1,000(360 + 300 − 2×328.6335), approximately 2,733 joules.

Direct contact produced no work and generated 8.30 J/K of entropy. The reversible comparison produces work with no entropy generation and ends in a different state. It does not claim that a real engine can attain the limit or that equal final temperatures alone make the endpoints identical. The path and transfers must remain visible.

Contents · Previous chapter · Next chapter

The endpoint formulas assume constant, appropriate heat capacities and equilibrium temperatures for each body. A material with strongly varying heat capacity needs a temperature-dependent integral. A phase transition needs an account that includes the relevant phase-change entropy and energy.

A container that warms, an enclosure that loses heat, or substantial work input changes the balance. Do not call every disagreement “entropy generation” while leaving missing energy transfers unexamined. Entropy generation must be inferred from a complete, appropriate entropy balance.

One thermometer reading may not represent an entire body. Internal temperature gradients, sensor lag and calibration can affect the evidence. A real validation would record where and when temperatures were measured, the thermal boundary and the uncertainty in capacities and readings.

For learning, supplied endpoint data are sufficient. There is no need to build a hot apparatus to understand the distinction. Explain which assumptions make a calculation meaningful, then identify which additional measurements a real system would require.

Contents · Previous chapter · Next chapter

Check one: a different common scale

Let each body have heat capacity 500 J/K, starting at 340 and 300 kelvin. Direct contact gives 320 kelvin. Each transfers 10,000 joules in magnitude. The combined entropy change is 500 ln(320/340) + 500 ln(320/300), approximately 1.96 J/K.

Check two: equal starting temperatures

Start both bodies at 320 kelvin with no other change. There is no initial thermal driving difference. The final temperature remains 320 kelvin, the energy transfers in this ideal equilibrium comparison are zero, and both logarithms are ln(1) = 0. The entropy-generation result is zero.

Check three: repair a temperature-unit error

A student substitutes Celsius values into ln(Tᶠ/Tⁱ). Ask whether converting the same physical temperatures to another arbitrary zero would change the result. It would, which exposes the mistake. Entropy temperature ratios require the absolute kelvin scale; ordinary temperature differences in an energy balance do not have the same issue.

Finish by explaining why a decreasing entropy for the hot body is compatible with the second law. The required answer names the combined isolated boundary and includes the larger increase of the cool body.

Contents · Previous chapter · Next chapter

Is entropy a form of energy?

No. Energy is measured in joules; entropy in joules per kelvin. Generating entropy does not create or destroy energy.

Can entropy decrease in one body?

Yes. The hot body in our example loses entropy. The second-law test uses the full boundary and its transfers, not an isolated sentence about one part.

Does insulation make a process reversible?

No. Insulation restricts heat crossing a boundary. Internal transfer across a finite temperature difference can remain irreversible.

Which reading route comes next?

For the foundation, return to The Gold Standard of Thermodynamics. For a fuller energy balance before adding entropy, use mixing temperatures when the container also warms. For performance ratios in another application, compare why an average COP can reverse a heat-pump comparison.

For energy usefulness and engineering context, continue with energy quality and exergy, thermal engineering and thermal energy storage engineering. The thermodynamic entropy terminology guide helps distinguish total, specific and generated entropy.

For the wider learning purpose, use The Gold Standard of Education and How X Works.

OpenStax University Physics Volume 2 provides the formal foundations in specific heat and calorimetry, reversible and irreversible processes and entropy. The numerical finite-body comparisons here are original applications of the explicitly stated constant-capacity model.

Contents · Previous chapter · Return to Thermodynamics