01Executive summary
Four milestones, two technologies, one shared constraint: both the gas turbine and the fission reactor are limited by materials, not by thermodynamics.
Frank Whittle ran an experimental turbojet in Britain in 1937; Hans von Ohain ran a comparable engine in Germany the same year, developed independently, and the Heinkel He 178 made the first turbojet flight in 1939. Chicago Pile-1 achieved the first self-sustaining fission chain reaction in 1942, and Calder Hall delivered nuclear electricity to a national grid at industrial scale in 1956. Both technologies have a straightforward efficiency argument that says run hotter, and both are held back by what the hot parts are made of.
02The gas turbine: continuous combustion
A piston engine does induction, compression, combustion and exhaust sequentially in one chamber. A gas turbine does all four simultaneously in separate components: air is compressed continuously, burned continuously, and expanded continuously through a turbine that drives the compressor, with the surplus available as thrust or shaft power.
The consequences are structural. There is no reciprocating mass, so the machine is smoothly balanced and can run at very high rotational speed. Power per unit mass and per unit frontal area are far higher than a piston engine of comparable output. Against that, the compressor and turbine must work continuously at conditions a piston engine sees only briefly, and part-load efficiency is poor because the compressor is designed for one operating point.
Turbine inlet temperature governs
Thermal efficiency and specific power both improve with hotter gas entering the turbine. That gas is hotter than the melting point of the blade alloys, so the entire discipline is about surviving it.
The compressor limits the map
Axial compressors operate close to aerodynamic stall. Too little flow for the pressure ratio and the machine surges — a violent flow reversal. Variable stators, bleed valves and multiple spools exist to manage this.
Everything is fatigue-limited
Blades experience high steady centrifugal stress plus vibratory excitation each time they pass a stator wake, at high temperature. Creep, thermal-mechanical fatigue and high-cycle fatigue all apply simultaneously.
Ingested damage is systemic
Birds, sand, ice and volcanic ash affect every stage. Certification includes ingestion testing because the failure of a single blade at speed is a containment problem.
Blade metal temperature is held below the gas temperature by three combined measures. Internal convective cooling passes compressor air through serpentine passages inside the blade. Film cooling bleeds that air through small holes to form an insulating layer over the surface. A ceramic thermal barrier coating adds further insulation. Directionally solidified and then single-crystal casting removes grain boundaries, which are where creep damage accumulates. The result is a component operating above the melting point of its own alloy — achieved not by finding a better material but by managing the thermal boundary condition. This is a general strategy: when a material limit blocks progress, change the environment the material experiences.
03Cycle choices and where each wins
| Configuration | Principle | Best suited to | Weakness |
|---|---|---|---|
| Turbojet | All useful output as jet thrust | High speed flight | Poor propulsive efficiency at low speed, high noise |
| Turbofan | Most flow bypasses the core, accelerated modestly by a large fan | Subsonic transport | Large frontal area, complexity of multiple spools |
| Turboprop and turboshaft | Nearly all output as shaft power | Low speed flight, helicopters, marine, mechanical drive | Propeller or rotor limits forward speed |
| Industrial gas turbine | Shaft power for generation or compression | Fast-start and peaking duty | Modest efficiency alone; poor at part load |
| Combined cycle | Gas turbine exhaust raises steam for a bottoming steam cycle | Baseload and mid-merit electricity | Capital cost, slower start, larger footprint |
| Steam cycle from a reactor | Nuclear heat raises steam | Steady baseload output | Lower steam temperature than fossil, so lower cycle efficiency |
Propulsive efficiency: the reason turbofans exist
Thrust equals mass flow times velocity change. The same thrust can be produced by giving a small mass of air a large acceleration or a large mass a small acceleration. The kinetic energy wasted in the jet varies with the square of that velocity change, so the large-mass, small-change option is far more efficient — provided the aircraft is slower than the jet. This is why high-bypass turbofans dominate airliners and why a turbojet, which does the opposite, is only sensible at high speed. The same reasoning applies to ship propellers and to helicopter rotor sizing.
04Fission: why a reactor is controllable at all
A fission event releases neutrons that can cause further fissions. The multiplication factor k is the average number of subsequent fissions caused by each fission. Below one the reaction dies away; above one it grows; at exactly one it is self-sustaining and steady.
The obvious concern is that neutrons travel at enormous speed, so a chain reaction should respond in microseconds — far faster than any mechanical control could act. That a reactor is controllable at all rests on a detail of fission physics that is easy to overlook.
A small fraction of fission neutrons — on the order of six or seven in a thousand for uranium-235 — are not released promptly but emitted seconds later by the decay of certain fission products. If the reactor is operated so that it is subcritical on prompt neutrons alone and reaches criticality only with the delayed contribution, then the rate of change of power is governed by those delay times, not by the prompt neutron lifetime. Control moves from microseconds to tens of seconds. Reactor operation is arranged entirely around staying within this delayed-critical regime, and the boundary at which prompt neutrons alone would sustain the reaction is the fundamental safety limit.
Feedback and inherent safety
Beyond control rods, a well-designed reactor is arranged so that physics opposes a power excursion without any action. If a rise in fuel temperature or a loss of coolant density reduces reactivity, then an increase in power is self-limiting. A negative temperature coefficient is therefore a design requirement rather than a convenient property, and this is the clearest available example of a general principle: build the corrective response into the physics rather than into a control system that can fail.
Why nuclear steam is less efficient than fossil steam
A coal or gas boiler can superheat steam to high temperature because the fire is external to the pressure boundary and the tube metal can be selected for it. A water-cooled reactor is limited by the coolant’s saturation conditions and by fuel cladding integrity, producing steam at substantially lower temperature. Lower peak temperature means a lower Carnot ceiling, and therefore lower thermal efficiency, regardless of how good the turbine is. It is a clean illustration that cycle efficiency is set by the source temperature the materials permit, not by the machinery.
05Takeaways for current practice
- When a material limit blocks progress, change its environment. Cooled, coated, single-crystal blades operate above their own melting point.
- Look for the physical property that makes control possible. Delayed neutrons convert an uncontrollable process into a manageable one; most controllable systems have an equivalent.
- Prefer inherent feedback to added control. A negative temperature coefficient cannot fail to actuate.
- Match the accelerated mass to the application. Propulsive efficiency reasoning generalises to any momentum-transfer problem.
- Peak cycle temperature is a materials decision. Efficiency arguments that ignore what the hot section is made of are incomplete.