Inside the J58.
Air. Fuel. Fire. Open the Blackbird’s engine and see why, past Mach 2, it stops behaving like a jet engine at all.
Why speed and altitude change thrust
Thrust is a subtraction. The engine is paid for the speed it adds to the air, not the speed the air leaves at: roughly mass flow × (exhaust speed − flight speed). At Mach 3 the air is already arriving at about 890 m/s, so the engine must hurl it out faster than that before it earns a single newton. That subtraction is the whole reason more Mach does not automatically mean more thrust — push the Mach slider up and watch net thrust fall even though nothing else changed.
What speed does buy you is free compression. The inlet converts the aircraft’s velocity straight into pressure — about 40 to 1 at Mach 3.2, before the compressor has done any work at all. That is why the thrust split swings so violently with speed: at Mach 2.2 it is 13% inlet, 73% engine, 14% ejector, and by Mach 3.2 it has become 54% inlet, 17% engine, 29% ejector (AEHS). The faster the aircraft goes, the more of the compression it gets for free from its own velocity, and the less of the work the engine itself is doing.
Altitude runs the other way. At the same Mach and throttle, lower air is denser, so more kilograms pass through the engine every second and thrust rises with them. But Mach 3 down low is not an SR-71 operating point: dynamic pressure and inlet heating climb past what the airframe and inlet will take. Equivalent airspeed (KEAS) is the number that expresses that load — it is the speed sea-level air would have to move to push on the aircraft this hard. Drop the altitude slider at Mach 3 and watch KEAS run away from you.
Past 500 KEAS this exhibit stops showing performance numbers, because they would be numbers for a flight condition the aircraft does not fly. That boundary is from NASA TM-104330, Figure 11. Clearing it does not mean a point is flyable: the real high-Mach climb schedule is narrower still, and inlet, engine and airframe limits all apply inside it. The sliders stay independent, and the recovery button only ever changes Mach — never your altitude or throttle.
The inlet profiles are reconstructed and its motion follows a nominal schedule. Pressure recovery is a generic estimate; shock positions, spillage, unstarts and the airframe ejector nozzle are not solved. Readouts describe the simplified engine cycle, not a validated installed SR-71 propulsion system.
Air is normally invisible; the blue tint reveals its motion. Pale mist is fuel spray. Blue-grey wisps reveal hot-gas motion through the exhaust even with reheat off; this tint is a teaching overlay, not smoke or flame. Motion is slowed; performance is estimated.
Your hand on the throttle.
Under the hood: station data, equations & assumptions
How to read this. Each station is a point along the gas path, and the two columns are what the gas carries when it arrives there. Move the throttle and watch the shape of it: pressure climbs through the compressor as shaft speed rises, the combustor adds a great deal of temperature while quietly losing a little pressure, and the turbine takes both back down again. That drop across the turbine is not waste — it is precisely the power being pulled out to drive the compressor.
Total (stagnation) values, not metal temperatures: what the gas would reach if brought to rest. All values below are model outputs.
| Station | Temperature | Pressure |
|---|
Gas speed and energy
| Passage | Axial speed | Total T | Total p |
|---|
Gas motion uses one shared 120× slowdown. The diffuser reduces velocity; staged heat release and dilution change mass flow and density inside the cans. The turbine extracts shaft work while total temperature and pressure fall. Local velocity can increase during expansion. Swirl and blade-row velocity triangles are not solved.
One linked thermodynamic cycle
Nothing here is set independently. Throttle sets fuel flow, fuel flow sets turbine-inlet temperature, that temperature sets the work available to the turbine, the turbine’s work sets shaft speed, and shaft speed sets the pressure ratio the compressor can produce — which changes how much air is flowing in the first place. Change one number and the loop settles somewhere new.
F = ṁexitVexit − ṁairVflight + (pexit − pambient)Aexit
In words: what leaves, minus what was swallowed, plus a bonus when the nozzle is still pushing harder than the air outside it. The middle term is the one people forget, and at Mach 3 it is enormous.
Assumptions: standard-atmosphere approximation; 8.8 reference compressor ratio (the 8.8:1 figure in the Smithsonian catalogue record below); 136 kg/s reference airflow; compressor efficiency 84%; turbine efficiency 90%; burner efficiency 98%; nozzle efficiency 95%; fuel heating value 43 MJ/kg; minimum idle fuel/air mass ratio 0.003 (assumed, not J58 calibration); maximum turbine-inlet target 1,360 K; maximum reheat target 2,100 K.
The efficiencies, temperature limits, flow normalization, bleed schedule and off-design relationships are educational assumptions. A 500 KEAS outer flight-condition screen and the assumed turbine-temperature limit reject unsupported performance readouts. No manufacturer maps, inlet-unstart model, complete engine/airframe limits, cooling network or validated transient dynamics are included. The screen does not simulate a fuel-control limiter or cap thrust at its sea-level rating. Out-of-range station data and momentum calculations below are hypothetical cycle extrapolations, not achievable engine performance. The main combustor retains idle fuel at 0% throttle; Shut down is the separate fuel cutoff. High inlet temperatures can make the idle floor exceed the assumed temperature limit, which is flagged as outside the useful model. Zero fuel commands report zero thrust; windmilling is excluded.
NASA · gas turbine cycle equations ↗One engine.
Two routes for the air.
Most air follows the engine's central path. At high flight speed, six external tubes tap the fourth compressor stage and carry that air straight to the afterburner, past stages five to nine, the combustor and the turbine. The main fuel control schedules the switch on compressor inlet temperature and engine speed.
It is worth doing because by then the inlet has already compressed the air about 40 to 1, and shoving air that hot through the back of the compressor costs more than it returns. That single piece of plumbing is why the J58 is called a turbo-ramjet: at Mach 2.2 the engine makes 73% of the thrust, and at Mach 3.2 it makes 17% while the inlet makes 54%.
Explore both routes, then click a part to see what it contributes.
Air follows the central path.