Let us name the one quantity we care about. ==Thrust F== is the forward push (in newtons) the engine gives the rocket. It is made of two shoves at the nozzle exit, and the picture below shows both.
Shove 1 — momentum thrust. In one second the engine throws out m˙ kilograms of gas at real speed ve. Throwing mass backward pushes the rocket forward (Newton's 3rd law), and the size of that push is
Fmom=m˙ve.
Shove 2 — pressure thrust. Look at the exit ring in the figure. The gas inside presses outward with pressure Pe; the surrounding air presses inward with Pa. Across the open exit area Ae there is no wall to cancel the leftover, so the net push is
Fpres=(Pe−Pa)Ae.
Add and repackage. Total thrust is F=m˙ve+(Pe−Pa)Ae. We define one make-believe speed c — the effective exhaust velocity — so that F=m˙c reproduces both shoves at once. Dividing the total by m˙:
c=ve+m˙(Pe−Pa)Ae
Symbol key for everything below: c effective exhaust velocity, vereal gas exit speed, Pe exit-plane gas pressure, Pa surrounding air pressure, Ae exit area, m˙ propellant leaving per second, F=m˙c thrust. The second figure previews the three expansion regimes the traps keep returning to.
False. Only when underexpanded (Pe>Pa). If overexpanded (Pe<Pa) the correction is negative and c<ve; if matched, c=ve.
TF2. "If the nozzle is optimally expanded, thrust becomes zero because the pressure term vanishes."
False. Only the pressure term vanishes; the momentum term m˙ve remains, so F=m˙ve=m˙c — usually the largest part of thrust.
TF3. "In a perfect vacuum a rocket makes no thrust because there is no air to push against."
False. Thrust comes from ejecting mass (Newton's 3rd law), not from pushing on air. In vacuum Pa=0, so in F=m˙ve+(Pe−Pa)Ae the pressure term becomes its maximum +PeAe, giving more thrust, not none.
TF4. "c stays constant as the rocket climbs, since ve, Ae and m˙ don't change."
False. Pa falls with altitude, so −PaAe/m˙ becomes less negative and crises during ascent — see Atmospheric Pressure vs Altitude.
TF5. "Because only the difference (Pe−Pa) matters, you may use gauge pressures for both."
True — if you are consistent. Gauge subtracts the same reference from each, so the difference is unchanged. The danger is mixing one absolute and one gauge, which is why the safe habit is absolute for both.
TF6. "Overexpansion gives lower thrust, so it is always bad and should never happen."
False. A nozzle optimised for high altitude is deliberately overexpanded at sea level; the small penalty low down is accepted to gain large vacuum performance. The real hazard is flow separation, not the sign alone — see Nozzle Expansion (Under/Over/Optimal).
TF7. "Isp and c carry different physics, so a change in one need not affect the other."
False. Isp=c/g0 with g0 a fixed constant; they are the same quantity in different units, so anything raising c raises Isp proportionally — see Specific Impulse.
TF8. "Tsiolkovsky's Δv=cln(m0/mf) should use ve, the real gas speed."
False. It uses the effective c, because c already bundles both momentum and pressure thrust into the number that actually accelerates the rocket — see Tsiolkovsky Rocket Equation.
SE1. "A student writes F=m˙ve and calls it the complete thrust equation."
Incomplete. The full Thrust Equation is F=m˙ve+(Pe−Pa)Ae; they dropped the pressure thrust, valid only when Pe=Pa.
SE2. "Since c=ve+(Pe−Pa)Ae/m˙, a bigger m˙ always makes c bigger."
Wrong. m˙ sits in the denominator of the correction, so a larger m˙ shrinks the pressure contribution and moves c toward ve; it does not scale c up.
SE3. "At sea level Pa=101.3 kPa and Pe=70 kPa, so the engine is underexpanded."
Error. Pe<Pa means overexpanded (exit pressure below ambient). Underexpanded is Pe>Pa. The verdict on the sign is correct only after comparing the two correctly.
SE4. "Pressure thrust exists because the outside air presses on the nozzle walls."
Misplaced. Wall pressures are already counted inside m˙ve (they shape the flow). The extra pressure thrust is the uncancelled net push (Pe−Pa) across the open exit ring, where there is no wall to balance it.
SE5. "The engineer sets Pe=Pa=0 to model a vacuum, giving pure momentum thrust c=ve."
Error. In vacuum Pa=0 but Pe is not zero — the gas still leaves with real static pressure. So the term is +PeAe/m˙>0 and c>ve, not equal.
SE6. "To boost thrust in space, just make the exit area Ae enormous with no other change."
Oversimplified. A bigger Ae raises the pressure-thrust term only while Pe stays positive; expanding the flow more actually lowersPe (and can cause overexpansion in atmosphere). You cannot inflate Ae in isolation — see Nozzle Expansion (Under/Over/Optimal).
WHY1. "Why do we invent c at all instead of keeping two separate terms?"
So every downstream formula — F=m˙c, Isp=c/g0, Δv=cln(m0/mf) — uses a single number, hiding the pressure bookkeeping inside one effective speed.
WHY2. "Why does the same engine have a higher Isp in vacuum than at sea level?"
Because c rises as Pa→0 (the back-push −PaAe/m˙ disappears), and Isp=c/g0, so higher c means higher Isp.
WHY3. "Why is momentum thrust independent of the surrounding air, but pressure thrust is not?"
Momentum thrust is the reaction to ejected mass (a Newton's-3rd-law effect internal to the exhaust), while pressure thrust is a difference against ambient Pa, so the environment directly enters it.
WHY4. "Why can overexpansion physically damage or destabilise the flow?"
With Pe<Pa the higher outside air pushes in and can peel the exhaust off the nozzle wall (flow separation), producing asymmetric, unsteady side-loads — a real structural risk, not just a thrust number.
WHY5. "Why does the pressure term use absolute and not gauge pressures in the standard statement?"
To avoid ambiguity: as long as both are absolute the difference is unmistakable. Mixing conventions silently shifts (Pe−Pa) by one atmosphere and corrupts c.
WHY6. "Why is 'perfectly expanded' the point of maximum thrust for a fixed geometry?"
At Pe=Pa the flow has converted the most pressure energy into velocity without the outside air pushing back; go further (overexpand) and ambient air steals thrust, so matched pressure is the peak for that nozzle at that altitude.
EC1. "What is c exactly at the single altitude where Pe=Pa?"
The correction is exactly zero, so c=ve. This happens at one specific "design altitude"; above and below it the term is nonzero.
EC2. "As altitude → orbit (Pa→0), what is the limiting value of c?"
It approaches its maximum cmax=ve+PeAe/m˙, the full vacuum effective velocity, with no ambient back-push left to subtract.
EC3. "If Pe=0 (gas expanded until its own pressure vanishes) in atmosphere, what happens?"
The correction becomes −PaAe/m˙<0, a pure penalty — a strongly overexpanded, physically unrealistic-in-atmosphere limit where ambient air alone pushes back; real flow would separate first.
EC4. "Can c ever be negative or zero?"
Only in a degenerate overexpanded extreme where the negative pressure term outweighs m˙ve; a working engine keeps c>0, since a non-positive c means no net forward thrust.
EC5. "What happens to the pressure term if m˙→0 while a small pressure imbalance remains?"
The term (Pe−Pa)Ae/m˙ blows up (divide by a tiny number), so c becomes ill-defined — a reminder that c is only meaningful while propellant is actually flowing.
EC6. "During throttle-down, m˙ drops but Pe often drops too — does c necessarily rise?"
Not necessarily. Smaller m˙ magnifies the correction, but a falling Pe shrinks (Pe−Pa); the net effect on c depends on which changes faster — you cannot judge from m˙ alone.