Opening the book…
A heat engine turns heat into useful work by running a working substance around a cyclic process: a closed loop that returns to its starting state each cycle. Because entropy and internal energy are state functions, both return to their starting values every cycle, so \(\Delta S = 0\) and \(\Delta E_{\rm int} = 0\) over a complete cycle. An engine draws heat \(Q_h\) from a hot reservoir, dumps waste heat \(Q_c\) to a cold reservoir, and delivers the difference as work (Figure 38.2):
Key Equation: Work of a heat engine
\[W_{\rm eng} = |Q_h| - |Q_c|. \tag{38.5}\]
The efficiency \(e\) is what you get out divided by what you put in: work delivered per unit heat drawn from the hot reservoir:
Key Equation: Thermal efficiency of a heat engine
\[e = \frac{W_{\rm eng}}{|Q_h|} = \frac{|Q_h| - |Q_c|}{|Q_h|} = 1 - \frac{|Q_c|}{|Q_h|}. \tag{38.6}\]
An efficiency of \(1\) (\(100\%\)) would mean no waste heat at all, \(|Q_c| = 0\), every joule of heat turned into work. The second law forbids it: such an engine would lower the entropy of the universe. Some heat must always be thrown away.
Engine Cycles on a \(P\)–\(V\) Diagram
In Chapter 37 we saw that the work a gas does during a process is the area under its path on a \(P\)–\(V\) diagram, \(W_{\rm by} = \int P\,dV\): positive when the gas expands and negative when it is compressed. An engine’s working substance goes around a closed loop, so in each cycle it expands along one part of the loop and is compressed along another. If the expansion happens at high pressure and the compression at low pressure, the gas does more work expanding than is done on it during compression. The net work is the area under the upper path minus the area under the lower path, which is simply the area enclosed by the loop (Figure 38.3):
Key Equation: Net work in a cycle
\[W_{\rm eng} = W_{\rm by,\,net} = {\text{area enclosed by the cycle on the }P\text{--}V\text{ diagram}}. \tag{38.7}\]
The direction matters. A loop traversed clockwise expands at high pressure and is compressed at low pressure, so the gas does positive net work: it is an engine. A loop traversed counterclockwise has net work done on the gas: it is a refrigerator or heat pump (Section 38.3). In either case \(\Delta E_{\rm int} = 0\) over the complete cycle, so the first law says the net heat absorbed equals the net work done by the gas, \(Q_{\rm net} = W_{\rm by,\,net}\).
To find an engine’s efficiency from its diagram we also need \(|Q_h|\), the heat taken in. This is the sum of \(Q\) over only those legs on which heat flows into the gas; the legs on which heat flows out make up \(|Q_c|\). The first law gives the heat on each leg as \(Q = \Delta E_{\rm int} - W = \Delta E_{\rm int} + W_{\rm by}\). For a monatomic ideal gas, \(E_{\rm int} = \tfrac32 nRT = \tfrac32 PV\), so the internal-energy change can be read straight off the diagram as \(\Delta E_{\rm int} = \tfrac32\,\Delta(PV)\).
Example 38.4: Efficiency of a rectangular cycle
A monatomic ideal gas is taken around the clockwise cycle of Figure 38.3: \(A\) (\(2.0\,\mathrm{L}\), \(1.0\times 10^{5}\,\mathrm{Pa}\)) \(\to\) \(B\) (\(2.0\,\mathrm{L}\), \(3.0\times 10^{5}\,\mathrm{Pa}\)) \(\to\) \(C\) (\(5.0\,\mathrm{L}\), \(3.0\times 10^{5}\,\mathrm{Pa}\)) \(\to\) \(D\) (\(5.0\,\mathrm{L}\), \(1.0\times 10^{5}\,\mathrm{Pa}\)) \(\to A\).
- Find the net work done by the gas per cycle.
- Find the heat absorbed per cycle, \(|Q_h|\), and the efficiency.
Solution
Work in SI units: \(1.0\,\mathrm{L} = 1.0\times 10^{-3}\,\mathrm{m^3}\), so the products \(PV\) at the corners are \(200\,\mathrm{J}\) at \(A\), \(600\,\mathrm{J}\) at \(B\), \(1500\,\mathrm{J}\) at \(C\), and \(500\,\mathrm{J}\) at \(D\).
(a) The loop is a rectangle, so the enclosed area is its width times its height:
\[W_{\rm eng} = (\Delta P)(\Delta V) = (2.0\times10^5\,\mathrm{Pa})(3.0\times10^{-3}\,\mathrm{m^3}) = \boxed{600\,\mathrm{J}}.\]
(b) Heat flows in on the two legs where \(PV\), and so the temperature, rises.
\(A \to B\) (isovolumetric): no work, so \(Q = \Delta E_{\rm int} = \tfrac32\,\Delta(PV) = \tfrac32(600 - 200)\,\mathrm{J} = 600\,\mathrm{J}\).
\(B \to C\) (isobaric): \(W_{\rm by} = P\,\Delta V = (3.0\times10^5)(3.0\times10^{-3}) = 900\,\mathrm{J}\) and \(\Delta E_{\rm int} = \tfrac32(1500 - 600)\,\mathrm{J} = 1350\,\mathrm{J}\), so \(Q = 1350 + 900 = 2250\,\mathrm{J}\).
Together these give \(|Q_h| = 600 + 2250 = 2850\,\mathrm{J}\), and
\[e = \frac{W_{\rm eng}}{|Q_h|} = \frac{600\,\mathrm{J}}{2850\,\mathrm{J}} = \boxed{0.21 = 21\%}.\]
(The other two legs give \(Q = -1500\,\mathrm{J}\) and \(-750\,\mathrm{J}\), so \(|Q_c| = 2250\,\mathrm{J}\), and \(|Q_h| - |Q_c| = 600\,\mathrm{J}\), as it must.)
Example 38.5: A cycle with a sloping leg
A monatomic ideal gas runs the clockwise triangular cycle shown: from \(A\) (\(1.0\,\mathrm{L}\), \(1.0\times 10^{5}\,\mathrm{Pa}\)) along a straight line to \(B\) (\(3.0\,\mathrm{L}\), \(3.0\times 10^{5}\,\mathrm{Pa}\)), straight down at constant volume to \(C\) (\(3.0\,\mathrm{L}\), \(1.0\times 10^{5}\,\mathrm{Pa}\)), and back to \(A\) at constant pressure.
- Find the net work done by the gas per cycle.
- Find \(Q\) for each leg, and the efficiency.
Solution
The products \(PV\) at the corners are \(100\,\mathrm{J}\) at \(A\), \(900\,\mathrm{J}\) at \(B\), and \(300\,\mathrm{J}\) at \(C\).
(a) The enclosed triangle has base \(\Delta V = 2.0\times 10^{-3}\,\mathrm{m^3}\) and height \(\Delta P = 2.0\times 10^{5}\,\mathrm{Pa}\):
\[W_{\rm eng} = \tfrac12(2.0\times10^{-3}\,\mathrm{m^3})(2.0\times10^5\,\mathrm{Pa}) = \boxed{200\,\mathrm{J}}.\]
(b) On each leg use \(W_{\rm by}\) = area under the path and \(\Delta E_{\rm int} = \tfrac32\,\Delta(PV)\), then \(Q = \Delta E_{\rm int} + W_{\rm by}\).
\(A \to B\) (sloping line): this leg has both work and a change in internal energy. The area under a straight line is a trapezoid, whose height is the average pressure:
\[\begin{aligned}W_{\rm by} &= \tfrac12(1.0 + 3.0)\times10^5\,\mathrm{Pa} \times 2.0\times10^{-3}\,\mathrm{m^3} = 400\,\mathrm{J}, \\ \Delta E_{\rm int} &= \tfrac32(900 - 100)\,\mathrm{J} = 1200\,\mathrm{J},\end{aligned}\]
so \(Q_{AB} = 1200 + 400 = +1600\,\mathrm{J}\).
\(B \to C\) (constant volume): \(W_{\rm by} = 0\) and \(\Delta E_{\rm int} = \tfrac32(300 - 900)\,\mathrm{J} = -900\,\mathrm{J}\), so \(Q_{BC} = -900\,\mathrm{J}\).
\(C \to A\) (constant pressure, compression): \(W_{\rm by} = (1.0\times10^5)(-2.0\times10^{-3}) = -200\,\mathrm{J}\) and \(\Delta E_{\rm int} = \tfrac32(100 - 300)\,\mathrm{J} = -300\,\mathrm{J}\), so \(Q_{CA} = -500\,\mathrm{J}\).
Heat enters only on \(A \to B\), so \(|Q_h| = 1600\,\mathrm{J}\) and
\[e = \frac{W_{\rm eng}}{|Q_h|} = \frac{200\,\mathrm{J}}{1600\,\mathrm{J}} = \boxed{0.125 = 12.5\%}.\]
Check: the three heats add to \(1600 - 900 - 500 = 200\,\mathrm{J}\), the net work, as the first law requires over a full cycle. Notice that most of the heat absorbed on the sloping leg (\(1200\,\mathrm{J}\) of the \(1600\,\mathrm{J}\)) goes into raising the gas’s internal energy, which it must then give back on the other two legs; only the enclosed area is left over as work.
Efficiency
How efficient can an engine be? Sadi Carnot answered this in 1824 by imagining the most efficient possible engine: one running a reversible cycle of two isothermal and two adiabatic steps, the Carnot cycle (Figure 38.4). He proved a remarkable thing: that the maximum efficiency depends on nothing but the temperatures of the two reservoirs. Written in terms of absolute temperature (a scale Kelvin introduced a generation later, partly on the strength of Carnot’s argument), it is
Key Equation: Carnot (maximum) efficiency
\[e_C = 1 - \frac{T_c}{T_h}, \tag{38.8}\]
with \(T_c\) and \(T_h\) in kelvin. No engine operating between two given temperatures can beat this; real engines, plagued by friction and irreversibility, fall short of it. The formula carries a sobering message: the only way to reach \(100\%\) efficiency would be a cold reservoir at absolute zero (\(T_c = 0\)), which is unattainable. To build a better engine you want the hottest possible source and the coldest possible sink.
Example 38.6: A steam engine’s ceiling
A steam engine boils water at \(500\,\mathrm{K}\) and exhausts to the outside air at about \(300\,\mathrm{K}\). What is its maximum possible thermal efficiency?
Solution
The best any engine can do between these reservoirs is the Carnot efficiency,
\[e_C = 1 - \frac{T_c}{T_h} = 1 - \frac{300\,\mathrm{K}}{500\,\mathrm{K}} = 1 - 0.60 = \boxed{0.40 = 40\%}.\]
Even a flawless, frictionless engine here must dump \(60\%\) of its intake heat as waste: a hard limit set by the temperatures alone, not by engineering skill.
Example 38.7: Pushing the hot side higher
The hottest temperature standard engine materials can withstand is about \(850\,\mathrm{K}\). With the atmosphere as the cold reservoir at \(300\,\mathrm{K}\), what is the highest theoretical efficiency?
Solution
\[e_C = 1 - \frac{T_c}{T_h} = 1 - \frac{300\,\mathrm{K}}{850\,\mathrm{K}} = 1 - 0.353 \approx \boxed{0.65 = 65\%}.\]
Raising the hot reservoir from \(500\,\mathrm{K}\) to \(850\,\mathrm{K}\) lifted the ceiling from \(40\%\) to \(65\%\), which is exactly why engineers chase ever-higher operating temperatures, limited in the end by what the materials can survive.
The Carnot limit also explains why the cycles of Section 38.3 perform so poorly. In the rectangular cycle of Example 38.4, the gas is coldest at \(A\) and hottest at \(C\), and since \(T = PV/nR\), \(T_C/T_A = 1500/200 = 7.5\). A Carnot engine working between those two temperatures would reach \(e_C = 1 - 1/7.5 = 87\%\), yet the rectangle manages only \(21\%\). The reason is where the heat flows. A Carnot engine takes in all of its heat at the hottest temperature and rejects all of it at the coldest. The rectangle takes in heat all the way up from \(T_A\) to \(T_C\) and rejects it all the way back down, so on average it absorbs heat well below its top temperature and discards it well above its bottom one, and every one of those heat flows across a temperature difference is irreversible.
Refrigerators
Run a heat engine backward and you get a refrigerator (or a heat pump). Instead of letting heat flow hot-to-cold and harvesting work, a refrigerator uses work input to force heat from a cold region to a hot one: the unnatural direction. This does not violate the second law: the work you supply (ultimately from the wall outlet) generates enough extra entropy to more than pay for the entropy decrease of moving heat the “wrong” way.
We rate a refrigerator not by efficiency but by its coefficient of performance (COP), the heat extracted from the cold space per unit work supplied:
Key Equation: Refrigerator coefficient of performance
\[K = \frac{|Q_c|}{W} = \frac{|Q_c|}{|Q_h| - |Q_c|}. \tag{38.9}\]
A good refrigerator has \(K > 1\): it moves more heat than the work it consumes, because it is pumping existing heat rather than creating it. As with engines, the second law sets a best-possible value, achieved by a reversible (Carnot) refrigerator,
Key Equation: Carnot (maximum) coefficient of performance
\[K_C = \frac{T_c}{T_h - T_c}, \tag{38.10}\]
which shows that a refrigerator works hardest, and its COP plummets, when it must maintain a large temperature difference between its cold interior and the warm room outside.
Checkpoint 38.3
(a) What is the maximum efficiency of an engine operating between reservoirs at \(500\,\mathrm{K}\) and \(300\,\mathrm{K}\)? (b) Another engine operates between \(600\,\mathrm{K}\) and \(400\,\mathrm{K}\): the same \(200\,\mathrm{K}\) difference. Is its maximum efficiency the same? (c) Why can no engine have an efficiency of \(1\)?
Answers to Checkpoints
Checkpoint 38.3. (a) The Carnot efficiency is \(e_C = 1 - T_c/T_h = 1 - 300/500 = 0.40\), or \(40\%\). (b) No: \(e_C = 1 - 400/600 = 0.33\), lower. The efficiency depends on the ratio of the temperatures, not their difference, and the same \(200\,\mathrm{K}\) is a smaller fraction of the hotter temperature. (c) An efficiency of \(1\) would require \(|Q_c| = 0\) (all the heat turned into work) or \(T_c = 0\). Dumping no heat would decrease the total entropy, since heat taken from the hot reservoir lowers its entropy and nothing compensates, violating the second law. And a cold reservoir at absolute zero is unattainable.
This page titled 38.3: Heat Engines was last modified on Thu, 12 Nov 2026 04:27:09 GMT and is shared under a CC BY 4.0 license and was authored, remixed, and/or curated by Michael Ragsdale.
The book could not be loaded just now.