The SSME low-pressure pumps (both fuel and oxidizer) are mounted to the vehicle structure and don't gimbal. The high-pressure pumps are gimbaling with the engine, so the propellant lines between low-pressure and high-pressure pumps must be flexible. That includes not only the pump discharge ducts (2-3 MPa), but also the turbine inlet (29 MPa) on both low-pressure pumps and turbine discharge (21 MPa) on the LPFTP. This looks a like a lot of articulating joints (3 per flexible line, based on illustration below) on pressurized lines, meaning a lot of potential failure points.

Some of the joints are visible in the picture below: enter image description here

What is the reason behind such engineering decision? Why are the low-pressure pumps not gimbaling with the engine? If they were mounted to the engine, all the lines downstream would be rigid. The flexibility of the propellant lines would be provided by the supply manifold upstream of the low-pressure turbopumps. This looks like a simpler solution, because of:

  • Lower pressure in these ducts (0.2 - 0.7 MPa)

  • Lower number of required flexible lines. I am not 100% sure by this, but it seems very likely, because there would be only 1 flexible line upstream of each low-pressure pump (2 for a single engine), instead of 3 downstream of the LPFTP + 2 after the LPOTP (that is 5 for a single engine). Multiply these 3 extra lines by the amount of articulating joints needed on each flexible line and it looks like a significant reduction of critical moving componens. There are of course other flexible lines connecting the engine with the orbiter (fuel tank pressurization, helium inlet, fuel bleed, etc.), but I think their number would be unaffected by the proposed change, so I am not counting them.

There must be some hidden reason why the SSME designers chose to fix the low-pressure turbopumps to the orbiter and forced themselves to use multiple articulating joints on high pressure lines. I would like to understand that reason.

Useful reference.

  • 2
    $\begingroup$ I have always wondered this myself. $\endgroup$ May 7, 2019 at 18:59
  • 1
    $\begingroup$ I wonder if the flex lines would be prone to cavitation at the low pressure turbopump inlet pressures. $\endgroup$
    – Tristan
    May 7, 2019 at 20:15
  • 3
    $\begingroup$ Speculating: Perhaps it has something to do with keeping the mass of what is gimbaled to a minimum, thereby reducing the amount of hydraulic power required for rapid gimbaling (which would be a concern during situations involving degraded hydraulics trying to cope with a demanding abort sequence, for example). $\endgroup$
    – Digger
    May 8, 2019 at 15:28
  • $\begingroup$ Related: space.stackexchange.com/q/21590/6944 $\endgroup$ May 10, 2019 at 16:34
  • 1
    $\begingroup$ You have to have a long articulated path on either the input or output of the LP pumps; you need one somewhere. Since cavitation at full flow was a concern (hence serial pumps), having the long path at suction pressure is not as good as having it at the LP pump output pressure. $\endgroup$ Jun 14, 2019 at 2:58

1 Answer 1


Start by thinking about why the low pressure pumps exist as all. Why can't the entire pumping task be done by single (multistage) pumps on the engine? Why did the engineers feel the need to have separated low pressure (LP) and high pressure (HP) pumps at all?

Because the engine gimbals, there needs to be a flexible lines to feed fuel and propellant to the engine yet still allow motion. That line also has to handle temperature (length) changes, work when iced up, have low mass, etc.

That line feeds large quantities of liquids near their boiling point to an HP pump. The pump is sucking really hard to move that mass. Because of the pressure drop of moving through that line, the end near the HP pump is below the pressure at the start of the line. If that line starts at just tank pressure, this is pretty much a recipe for cavitation(i.e. spontaneous boiling), which causes all sorts of problems up to destruction.

To avoid cavitation, the LP pumps were included. Pretty much by definition, they had to be before the long flexible line so that their output would pressurize that line; the additional pressure from the LP pump meant that that the input suction pressure of the HP pumps would never be below the vapor pressure of the fluids, so they wouldn't cavitate (i.e. spontaneously boil). So the LP pumps were mounted to the large, fixed tank piping, hence directly to the Shuttle framework.

TLDR: The LP pumps entire job was to pressurize the feed to the HP pumps. That feed starts at the bottom of the tank piping, so that's where they were put.

(Another approach to cavitation reduction is subcooling the fluids, which is interesting on its own)

  • $\begingroup$ Thank you. Would it be possible to share more technical details? For example: does it mean that having the articulated path upstream of LP pump would increase the risk of cavitation? How exactly? The pressure and flow rate would be unaffected. Also, why the long articulated path is the only option here? There exist other alternatives of flexible lines, as seen for example here: ui.adsabs.harvard.edu/abs/1977NASSP8123....../abstract $\endgroup$
    – mpv
    Jun 14, 2019 at 11:46
  • $\begingroup$ Note that the SSME design required at least two additional flex lines - for the low pressure turbine drive fluids - that wouldn't have been required if the low pressure pumps gimbaled with the engine. $\endgroup$ Jun 14, 2019 at 12:13
  • $\begingroup$ I've rewritten the answer to try make it a bit clearer.The pressure would not be unaffected: Pumps exist to make the pressure after them more than the pressure before them. $\endgroup$ Jun 14, 2019 at 12:51
  • 2
    $\begingroup$ @OrganicMarble Agree that having separate LP pumps is complex. They really must have needed a reason for all that, because those engineers were pretty motivated to reduce mass and complexity. The NASA press kit (hardly engineering documentation) says it was to prevent cavitation. Perhaps they found that a non-cavitating line had to be bigger, heavier and more of a risk. LH2 and LOX are really motivated to boil... $\endgroup$ Jun 14, 2019 at 12:58
  • $\begingroup$ Hmm, there was a whole lot of plumbing before the low pressure pump inlet. Not sure you're making the case here. $\endgroup$ Jun 14, 2019 at 13:03

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.