Lesson 13 of 14

Steering, Plumes & Testing

Thrust vector control — gimbals, flexible nozzles, vanes and side injection — and why a rocket is an inverted pendulum; what a plume's shock diamonds, afterburning and smoke reveal; and how engines are proven on the test stand.

5 min3 live simulations
you are here
The question

A launch vehicle is a tall, thin tube pushed from the bottom by a flame, climbing through air that is trying to flip it over. How do you steer a flame — and what can you learn from looking at one?

Thrust vector control

A rocket has no rudder that works in thin air or in space, and its engine pushes from behind the centre of mass. Steering means tilting the thrust, so that it passes a little to one side of the centre of mass and creates a turning moment. Sutton surveys the ways to do it, collectively called Thrust vector controlAny means of steering a rocket by tilting its thrust: gimbals, flexible nozzle joints, jet vanes, or side injection.:

  • Gimballed engines. The whole engine pivots on a GimbalA pivot that lets the whole engine swivel a few degrees so the thrust line passes off the centre of mass, producing a turning moment., moved by hydraulic or electric actuators — typically by up to 5–10°. Almost every liquid engine is steered this way.
  • Flexible nozzle joints. Solid motors can't tilt their whole chamber, so the nozzle alone swivels on a flexible bearing of laminated rubber and metal shims, as on the Shuttle boosters.
  • Jet vanes. Heat-resistant vanes in the exhaust deflect the jet, as on the V-2. Simple, but they cost thrust and erode.
  • Secondary injection. Injecting a liquid or gas through the side of the nozzle creates an oblique shock that deflects the jet: no moving parts in the exhaust.
  • Differential throttling of multiple engines, and small reaction control thrusters for roll control and for steering in space.

A gimbal of just a few degrees is enough, because the moment arm — engine to centre of mass — is long, and because a well-flown rocket needs only small corrections. The difficulty is that it needs them constantly.

SteerBalance a rocket on its own flame
gusty
Tilt0.0°
Gimbal0.0° / ±6°
Worst tilt0.0°
Challenges0/3
  • Fly by hand through max-Q (30 s) in gusty wind.
  • Tune the autopilot to survive a storm (wind ≥ 0.8).
  • Fly the autopilot through gusts with the tilt never exceeding 2°.
A rocket pushed from the bottom is an inverted pendulum, and in thick air its centre of pressure sits ahead of its centre of mass, so the air tries to flip it. Swivelling the engine a few degrees on its gimbal moves the thrust line off the centre of mass and pushes back. Fly it by hand (hold the buttons or the arrow keys) — then try the autopilot.

Aerodynamically, most launch vehicles are unstable: without big tail fins, their centre of pressure lies ahead of their centre of mass, so any angle of attack produces a moment that increases it. Near max-Q that overturning moment is at its largest. The control system must sense attitude and rotation rate and move the gimbals fast enough to catch every disturbance. The autopilot in the simulator is a classic PD controller: proportional gain supplies the restoring stiffness, derivative gain the damping. Real vehicles add filtering, gain scheduling with dynamic pressure, and compensation for fuel sloshing and structural bending.

Estimate first
The Saturn V's F-1 engines could gimbal. What was the maximum gimbal angle, in degrees?
0.10110
Drag along the line to place your guess — each tick is ten times the last.

Reading a plume

The exhaust plume is the last part of the propulsion system, and it is full of information. Sutton's chapter on plumes covers their shape, their radiation, their noise, their chemistry — and their effect on the vehicle.

ReadReading a plume
0.0 km
16
Exit / ambient pressure0.67
Fully expanded Mach3.4
Jet width / nozzle exit0.9×
Afterburningyesfuel-rich exhaust meets air
Challenges0/3
  • Find the altitude where this nozzle is perfectly expanded.
  • Make a Mach disk appear.
  • Make the plume more than five times wider than the nozzle.
A rocket plume is a map of the pressure mismatch at the nozzle exit. Over-expanded, the air squeezes the jet and oblique shocks cross inside it; under-expanded, the jet bulges outward and reflects expansion waves. Either way the jet oscillates around its 'fully expanded' diameter in a train of shock cells — the diamonds — whose spacing grows as the air thins.

What you can read from it:

  • Shock diamonds. A jet that is not perfectly expanded over- or under-shoots the ambient pressure and oscillates around it, in a train of expansion waves and shocks. The bright spots where shocks cross are heated gas: the Mach diskA strong normal shock in an over- or under-expanded jet; together with the oblique shocks it forms the bright “shock diamonds”., with a Mach disk — a normal shock — when the mismatch is large. Their spacing grows as the air thins.
  • Plume blooming. High up, the exhaust leaves far above ambient pressure and expands enormously. Upper-stage plumes can be many times wider than the vehicle.
  • Colour and luminosity. Hydrogen–oxygen exhaust is nearly transparent: water vapour glowing faintly blue and pink. Kerosene exhaust is bright orange-yellow from glowing soot. Aluminized solids glow white-hot with oxide particles and leave a dense white trail.
  • Afterburning. Fuel-rich exhaust still contains hydrogen and carbon monoxide. When it meets air, it burns again: the plume brightens and heats, and it releases more energy than the rocket used for thrust.

Plumes matter to engineers for practical reasons, too: they heat the vehicle's base, contaminate optics, interfere with radio links, generate intense acoustic loads at launch, and produce the signature that tracking sensors watch.

Proving it on the ground

No rocket engine flies without being tested, often hundreds of times. Sutton's chapter on testing describes a hierarchy: component tests (injectors, pumps, valves), engine firings on a test stand, integrated stage tests, and finally flight tests, supported by facilities that can simulate altitude and absorb the exhaust.

A Static fireA test firing of an engine or vehicle held down on a stand, instrumented to measure thrust, pressures, flows and temperatures. holds the engine down and measures everything: thrust from load cells, chamber and feed pressures, temperatures, flows, vibrations. From them engineers compute the quantities we have used all course — Isp, c∗c^* efficiency, CFC_F — and compare them against predictions. Transient behaviour at start-up and shutdown gets as much attention as steady running, because that is where most engines fail.

MeasureStatic fire: from wiggly traces to Isp
05101520050100time (s) — drag to set the averaging windowthrust (kN)
0510152080010001200tank mass (kg)
Average thrust84.2 kN
Mass flow33.42 kg/sslope of the tank mass
Isp (sea level)257.0 svacuum-corrected 290 s
c* efficiency85.2%c* = p_c A_t / ṁ = 1534 m/s
Your window includes transients. During start-up and shutdown thrust and mass flow are out of step, and they bias the average — typically downward.
Challenges0/2
  • Measure Isp to within 0.5% of the true value, choosing the window by hand.
  • Read off the injector’s c* efficiency.
A 100 kN engine held down on a test stand for 14 seconds. A load cell measures thrust; weigh-cells under the propellant tanks measure how fast mass leaves. Drag the edges of the shaded window to choose which part of the run to average. Include the start-up or shutdown transients and your Isp is wrong.
Check your intuition
A test team measures thrust and propellant flow at sea level and computes an Isp of 290 s. The same engine will fly on an upper stage in vacuum. Can they predict its vacuum Isp from the same test?
Recap
What to carry forward
  • Thrust vector control tilts the thrust off the centre of mass: gimbals, flexible nozzles, jet vanes, secondary injection, differential throttling, RCS.
  • An aerodynamically unstable rocket is an inverted pendulum; a feedback controller with stiffness (P) and damping (D) keeps it upright through max-Q.
  • Plumes show the pressure mismatch as shock diamonds and Mach disks, bloom at altitude, and may afterburn in air.
  • Testing proceeds from components to full vehicles; static fires yield Isp, c∗c^* and CFC_F — but only from steady-state data, and with care over transients.

Next, the finale: pick a mission and design the propulsion system that flies it.

Finished this lesson?Mark it complete to track your progress through the course (saved in this browser).
Up next · Lesson 14Mission Design StudioCan you design a propulsion system that flies every mission?Keep going →