Index Home About Blog
From: (Henry Spencer)
Subject: Re: LH2 coolant for SRB or F-1 rocket nozzels -- What drives the 
Date: Wed, 31 May 2000 23:14:30 GMT

In article <xdZY4.74341$>,
lament  <> wrote:
>I believe that centrifugal pumps are used to deliver fuel and LOX into
>the chamber, and that the pressure in the reaction chamber is on the
>order of 2,000 psi.

LH2 pumps are sometimes axial, but not always, and everything else is
usually pumped with centrifugal pumps.  Chamber pressure varies a lot,
depending on the engine requirements and how aggressive the designer is
being.  Older engines often run at modest pressures, e.g. 475psi for the
RL10, but modern ones tend to run somewhat higher.

>The flow rate is immense. What is used to drive the turbines that pump
>LOX and fuel into a rocket chamber?

There are various methods -- "engine cycles" -- but the usual general
approach is to burn some of the fuel and oxidizer in a separate
"preburner" or "gas generator" to provide turbine gas.  What happens to
the turbine exhaust varies; sometimes it's dumped overboard, sometimes it
goes into the combustion chamber.

>Saturn 5 had five rocket engines, each generating 1.5 million pounds of
>thrust. But there seems to be no other exhaust visible except for the 5

Each F-1 engine had its own pumps and pump turbine, with the turbine
exhaust dumped.  However, the dumping is not obvious... because it went
into the main exhaust nozzle, midway down (this is what the big tapered
tube encircling the nozzle is for).  This not only solved the awkward
problem of how to dump the hot turbine gas without having it go places you
don't want, but it also helped protect the nozzle wall against the
much-hotter main exhaust.
Microsoft shouldn't be broken up.       |  Henry Spencer
It should be shut down.  -- Phil Agre   |      (aka

From: (Henry Spencer)
Subject: Re: Turbopump questions
Date: Sat, 16 Sep 2000 07:14:23 GMT

In article <>,
Greg D. Moore <> wrote:
>1) you're often dealing with cyrogenic fuels, which means they get cold
>and brittle.  And different pieces will shrink different amounts due to
>the coldness.

While this adds extra problems, they are well understood.  Commercial
cryogenic pumps have an operating life of decades, like most other heavy
industrial gear.

>2) They have to be extremely effecient and not very massive.  This means
>moving a LOT of liquid with as lightweight equipment as possible.
>This leads to designs with very tight tolerances.

There is *pressure* in that direction, but one can resist it.  Management
and design philosophy, rather than technical necessities, are the key
issues here.

The RL10's turbopumps seem to have an almost unlimited life, including
long runs and multiple restarts, demonstrated both in space and on DC-X.
They were designed with somewhat greater margins than is usual in the
rocket business, partly because the engine started out to be a bit more
powerful than it ended up, partly because they were designed by aircraft
people rather than rocket people.  This did not stop the RL10 from being
quite a high-performance engine.
Microsoft shouldn't be broken up.       |  Henry Spencer
It should be shut down.  -- Phil Agre   |      (aka

From: Bruce Dunn <>
Subject: Re: Turbopump Performance
Date: Fri, 22 Sep 2000 22:06:20 GMT

Some feeling for the extent to which turbopump problems have
historically plagued the development of rocket engines can be
appreciated from US experience related in the book:

 "History of Liquid Rocket Engine Development in the United States
1955-1980"  Stephen E. Doyle, Editor :  AAS History Series, Volume 13 -
Proceedings of an AAS History Colloquium; American Astronautical
Society, San Diego, 1992.

roughly totaling the problems listed below gives the following numbers,
from which it is clear that over half the problems in engine development
programs described in the book were related to turbopumps:

turbopump              20
regenerative cooling   3
combustion             4
valves                 2
miscellaneous          5

Developmental problems in liquid rocket engines:

LR-87 (Titan 1 first stage) and LR-91 (Titan 1 second stage);  LOX and

- cooling jacket sizing problem leading to unsatisfactory propellant
phasing during starts
- problems with turbopump starting
- igniter damage to regenerative cooled chamber walls
- cokeing of turbopump turbine nozzles

LR-87 (Titan 2 first stage) and LR-91 (Titan 2 second stage);  Aerozine
50 and N2O4
(LOX/RP-1 version extensively redesigned for storable propellants)

- combustion instability
- "soft goods" [seals, O-rings] deteriorating with time in stored
- marginal turbopump gearbox load capacity, requiring redesign of shaft
and bearings
-  turbopump gearbox overheating, requiring redesign of lube oil cooling
and pumps

MB-1 and MB-3 (Thor first stage); LOX and RP-1

- turbopump geartrain problems involving lubrication and axial shaft
- turbopump turbine blade cracking
- combustion instability

H-1 (Saturn 1 first stage); LOX and RP-1
RS-27 (Delta first stage); LOX and RP-1

- these are refinements or derivatives of the MB-3, and no specific
development problems are noted

MA-1, MA-2, MA-3, MA-5 (Atlas first stage); LOX and RP-1

- turbopump geartrain lubrication problems
- turbopump turbine blade failures
- combustion instability
- LOX turbopump shaft deflection leading to rubbing followed by LOX pump

RL-10 (Centaur upper stage); LOX and LH2

- problems in using oil lubricated gearboxes in turbopumps using
cryogenic fluids (solved by using dry gears cooled by hydrogen)
- gearbox lockup due to diffusion bonding of dry gears in a hydrogen
atmosphere (solved by switching gear materials and adding a dry
- burnthrough of regeneratively cooled combustion chamber
- inadequate combustion efficiency (solved by combustion chamber
- thermally induced warping of injector plates
- unreliable ignition
- thrust overshoot (overspeeding of turbopumps during startup)
- difficulties in cooling down engine prior to ignition (solved by a
pre-launch cooldown using liquid helium)
- oxygen boost pump failure due to moisture contamination during LOX
loading (solved by spinning the boost pumps with a cold gas system
during LOX loading)

Space Shuttle Main Engine; LOX and LH2

- start and shutdown sequences (the engine has an exceedingly complex
propellant system, and the exact sequencing of valves is critical to
allow startup without exceeding temperature or pressure limits in
critical systems; developing a workable startup sequence required 35
weeks, 37 tests, and 13 replacement turbopumps)

- high pressure fuel turbopump subsynchronous whirl (rotor gyration
ending in rubbing and bearing failure)

- high pressure oxidizer turbopump explosions (various causes, including
LOX seal leaks,  vibration followed by fires, and a sensor problem which
led to a badly fuel rich turbine gas which was so cold that the water in
it froze, causing turbine rubbing, failure, and ignition)

- high pressure fuel turbopump blade failures

- main oxidizer valve fire (vibration at a resonant frequency leading to
bolt loosening, further vibration, and ignition of parts of the valve)

- main fuel valve fracture (cause uncertain, but probably due to a
highly stressed valve designed of titanium with inadequate margins -
fixed by redesigning to try to eliminate all possible problems which
could have caused the observed failure)

- nozzle feed line fractures ( vibration induced fatigue, and welds in
Inconel 718 made incorrectly with welding filler wire intended for
Inconel 62, which has half the strength; a reinspection of existing
fabricated parts revealed that 3359 welds out of 16,360 were made with
the wrong wire !)

- fuel preburner burnthroughs (oxygen trapped in an acoustic cavity
eroded the metal, an independent problem caused by a warped preburner
face plate, and poor flow in the unit fixed by various redesigns)

From: Bruce Dunn <>
Subject: Re: How much power does a rocket lose to its turbopump?
Date: Thu, 28 Sep 2000 02:44:47 GMT

toby wrote:
> If the shuttle engines generate 100X watts of power in total.
> How much of this is lost to the turbopump?

In the space shuttle main engine, essentially none of the power is
"lost" in the turbopumps.  All mass and all energy from combustion (even
in the preburners) ends up in the combustion chamber, where it
contibutes to the thrust.

Gas generator cycle engines however burn propellant at non-optiumum
mixture ratios, and exhaust the working gas without generating much
thrust.  Moderate pressure LOX/hydrocarbon gas generator engines ( F1
from the Saturn, RS-27 from the Delta) burn roughly 2% of the propellant
mass flow in the gas generator.

From: "Paul F. Dietz" <>
Subject: Re: 2 fuels, 1 oxydizer SSTO...
Date: Wed, 13 Dec 2000 21:06:36 -0600

Henry Spencer wrote:

> >...My question is, would it also be feasible to have a
> >kerosene and liquid hydrogene turbopump?
> No, because the volume to be pumped is so much higher with hydrogen, and
> volume is what mostly drives pump design.

Also, because the pressure achieved by a centrifugal pump
is proportional to the density of the liquid being
pumped.  You wouldn't want to use the same pump
on two liquid of radically different density, while
at the same time injecting the oxidizer at a fixed



Index Home About Blog