Falcon 9 first stage use turbopumps as well as helium pressurisation system. Isn't the turbopumps enough to pump in the propellants to the combustion chamber?

  • $\begingroup$ Generally a certain pressure is required at the inlets to prevent cavitation of the pumps. $\endgroup$ Oct 12, 2020 at 18:50
  • $\begingroup$ @Organic Marble, Wouldn't the RP1 and LOX self pressurise to allow safe start of the turbopumps? $\endgroup$
    – Ashvin
    Oct 12, 2020 at 18:55
  • $\begingroup$ Oxygen can autogenously pressurize from boiloff or from tapping off heated LOX from the engine, but RP1 can't and needs an inert pressurant. $\endgroup$
    – Tristan
    Oct 12, 2020 at 18:59
  • 1
    $\begingroup$ Using turbopumps without tank pressurisation would destroy both the tank and the pump. The tank pressure must be higher than ambient pressure to avoid crumpeling. A too small pump input pressure would cause cavitation. $\endgroup$
    – Uwe
    Oct 12, 2020 at 19:49

2 Answers 2


Pumps are great at creating positive pressure, hundreds of bars. But with tank at ~1 bar, atmospheric pressure, they can only create a grand total of 1 bar of suction - can't go more vacuum than vacuum, can't create a negative density, can only go from original to zero - and 1 bar of pressure is a pretty meager amount, when this kind of flows is involved. And without a gas to replace the liquid already pumped out, the pressure would be dropping even more.

And not only does it limit the supply - As demand exceeds supply, vacuum bubbles start forming - aka cavitation, areas where new liquid can't flow in fast enough. And cavitation is bad. Liquids are very badly compressible, and such a vacuum bubble, when the flow catches up, vanishes completely as the liquids on its sides meet up - and does so quite violently. It causes a pressure wave that travels through the liquid at speed of sound (in that liquid) and can cause trouble wherever it hits, be it sending tank walls into vibration that can damage bearings, be it impacting turbine blades, bending them, or facilitating more cavitation where conditions were "almost but not quite there" and multiplying itself - never mind pump blades churning vacuum don't do anything useful, a lot of energy wasted on spinning pointlessly moving way less liquid than they could - if that liquid was there.

So you need that extra pressure to overcome the resistance of the plumbing, valves, acceleration of the liquid entering the pipes etc, so that the turbopumps never cause cavitation.

  • 3
    $\begingroup$ In other contexts one would call this approach push-pull, as the over-pressure from the helium pushes the liquid out of the tank, and the suction from the pumps pulls the liquid towards the pumps. No idea whether this is used in this context, however, I find push-pull describes the result rather good. $\endgroup$
    – Dohn Joe
    Oct 13, 2020 at 7:33
  • $\begingroup$ @DohnJoe In pumps, a positive pressure at the inlet is called 'hydraulic head' and is typically expressed in meters (although probably not in space applications), and it is implicitly understood that a higher head is better (think of a submersible pump). I would personally associate push-pull with bidirectionality (like in a push-pull amplifier), not with increasing hydraulic head. $\endgroup$
    – Sanchises
    Oct 13, 2020 at 9:08

In a turbopump feed system it is still necessary to pressurize the tanks slightly (10 to 50 lb/in2) in order to prevent pump cavitation.

Rocket Propulsion Elements, Sutton, 4th edition, p. 223

Different vehicles have used stored onboard gas (typically helium) or autogenous propellants to provide this pressurization.

Further reading:

  • 3
    $\begingroup$ Crazy how a scientific/engineering book from 2001 using lb/in^2 $\endgroup$
    – Michael
    Oct 13, 2020 at 6:33
  • 1
    $\begingroup$ @Michael Rocket science is a bit of a misnomer. A better term is "rocket engineering". Use of customary units remains very widespread to this day amongst engineers in the US. I have worked for / consulted for multiple new space endeavors who explicitly start with the goal of using metric throughout rather than customary units. While I laud that goal, it never seems to work, at least not completely. $\endgroup$ Oct 13, 2020 at 7:37
  • 1
    $\begingroup$ @Michael Imperial units were still in use in 1998. NASA institutionally may have learnt from that, but individual engineers are likely more resistant to change. $\endgroup$
    – Graham
    Oct 13, 2020 at 8:20
  • $\begingroup$ @Michael my copy is from the 1970s. $\endgroup$ Oct 13, 2020 at 11:45
  • $\begingroup$ @Graham the shuttle program ended in 2011, and was all English units (not Imperial). $\endgroup$ Oct 13, 2020 at 11:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.