Centaur-2A upper stage of an Atlas
is a rocket
designed for use as the upper stage of space launch vehicles
. Centaur boosts its satellite
payload to its final orbit
or, in the case of an interplanetary space probe
, to escape velocity
. Centaur was the world's
first high-energy upper stage, burning liquid hydrogen (LH2
) and liquid oxygen (LOX
Centaur, named after the centaurs
, was the brain child
of Karel J. "Charlie" Bossart
(the man behind the Atlas
ICBM) and Dr. Krafft A. Ehricke
, both Convair
employees. Their design was essentially a
smaller version of the Atlas concept, using lightweight stainless steel
"balloon tanks" whose
structural strength was provided by the pressure of the fuel
Centaur uses an ingenious common double-bulkhead to separate the
LOX and LH2 tanks. The two stainless steel skins are separated by a
0.25 inch (6.4 mm) layer of fiberglass. The extreme cold
of the LH2 on one side creates a vacuum within the fiberglass
layer, giving the bulkhead a low thermal conductivity, and thus
preventing heat transfer from the relatively warm LOX to the super
cold LH2. It is powered by one or two RL10
rocket engines (SEC and DEC variants respectively).
Development started in 1956 at NASA's
Center, now the Glenn Research Center, but proceeded slowly, with the first
(unsuccessful) test flight in May 1962.
In the late 1950s
and early 1960s Centaur was proposed as a high energy upper stage
for the Saturn I
, Saturn IB
and Saturn V
rockets, under the designation S-V
("Saturn V") in
accordance with the numbering of other stages of Saturn rockets.
However, the first successful Centaur flight did not take place
until 1965, by which point NASA had replaced the Centaur with much
larger upper stages on their designs.
From 1966 to 1989, the Centaur-D was used as the upper stage for 63
Atlas rocket launches. 55 of these launches were successful.
From 1974 to 1977, the Centaur-D-1T was used as the third stage on
7 Titan IIIE
launches, 6 of which were
successful. Spacecraft launched by these vehicles included Viking 1
, Viking 2
, and Voyager 2
A major change to the Centaur occurred in the early 1980s with the
removal of the hydrogen peroxide powered boost pumps and attitude
control system from the vehicle. Instead the RL-10 engines were fed
directly via tank pressure- resulting in significant reduction in
system complexity. A hydrazine monopropellant attitude control
system replaced the previous hydrogen peroxide system.
A new version, the Centaur-G, was developed for use with the
but was never used due
to tougher safety rules imposed after the Challenger
. This Shuttle/Centaur configuration changed the
hydrogen tank diameter to 14 feet while retaining the 10 foot
diameter oxygen tank. The geometry was optimized for installation
into the Space Shuttle Orbiter payload bay. Its initial mission was
to be the Galileo scientific probe to Jupiter. The Centaur systems
were dwarfed in their complexity by the supporting fluids, avionic
and structural systems which were integrated into the Centaur
Integrated Support System or CISS. In addition to more mundane
tasks these systems were required to rapidly dump propellants
overboard in the event of a Return
To Launch Site
(RTLS) abort. This was required to permit the
orbiter to land safely. These contingency, emergency and abort
provisions effectively amplified system complexity to an extreme
level and drove the majority of the systems design.
The decision to terminate the Shuttle/Centaur program spurred the
US Air Force
to create the Titan IV
, which used a similar Centaur-T with a 14
foot hydrogen tank diameter as its final stage. This vehicle was
capable of launching payloads which had originally been designed
for the Shuttle-Centaur combination. In the Titan 401A
configuration, a Centaur-T was launched 9 times between 1994 and
1998. In the Titan 401B configuration, a Centaur-T was launched 7
times, with one failure, between 1997 and 2003. The last
Titan-Centaur launch was in 2003. The 14 ft diameter Centaur
design has now been effectively retired.
Another major reconfiguration was done for the Atlas III vehicle
with a change from dual RL-10 engines as standard to a single
RL-10. This change was accomplished while retaining the ability to
revert to dual engines should mission requirements dictate. For
most missions however a single RL-10 is optimal or adequate and
hence a substantial reliability and cost benefit was
A "Common Centaur" was unveiled by LM on November 30, 1999. The
stretched stage was 38.5 feet (11.68 m) in length, 5.5 feet (1.7 m)
longer than the Centaur used on Atlas IIA and IIAS rockets at that
Although Centaur has a long and successful history in planetary
exploration, it has had its share of problems, especially early on:
- May 8, 1962: Centaur weather shield separated early; stage
- June 30, 1964: RL-10 hydraulic actuator pump shaft broke,
preventing one of the two RL-10 engines from vectoring. This lead
to no roll control, and uncovering of LOX inlet.
- December 11, 1964: Restart attempt failed, due to problem with
- April 7, 1966: Centaur did not restart after coast - ullage
motors ran out of fuel.
- August 10, 1968: Centaur restart failed.
- May 9, 1971; Centaur guidance failed, destroying itself and the
Mariner 8 spacecraft bound for Mars orbit.
- Feb 11, 1974: Titan-Centaur; boost pump failed.
- June 9, 1984: Centaur LOX tank failed; no restart.
- April 18, 1991: Centaur failed due to icing of hydrogen pump
impellor blades (not understood at the time)
- August 22, 1992: Centaur failed to restart (icing problem
- April 30, 1999: Launch of the USA-143 (Milstar-3) comsat failed when a Centaur programming
error resulted in incorrect burn times, placing the satellite in a
15, 2007: the engine in the Centaur upper stage of an Atlas V shut down early, leaving its payload—a pair
Reconnaissance Office ocean surveillance
satellite -- in a lower than intended orbit. The failure
was called "A major disappointment", though later statements claim
the spacecraft will still be able to complete their mission. The
cause was traced to a stuck-open valve that depleted some of the
hydrogen fuel, resulting in the second burn terminating 4 seconds
early. The problem was fixed and the next flight was nominal.
, derivatives of the 10 foot diameter Centaur-3, with either one or two RL-10A4-2 engines, continue to be used as the upper stage of the Atlas V EELV rocket, the successor of the Titan-Centaur configuration.
There is a possible use of the Centaur as an upper stage on the new
Delta IV Heavy
which started test flights in 2004, and may even be used as a
high-energy "kick motor" for planetary probes launched onboard the
125-ton Ares V
, which will see its first
flight around 2018..
A Centaur has also been proposed as an upper stage for the proposed
Ares-V alternative, the Jupiter-232
. It would
be used on the Jupiter until the J-2X
upper stage engine became
The Centaur has a planned evolutionary upgrade which changes the
tank diameter to 5.4m and increases propellant load from 1.5 to 6.0
times that of the present Atlas V configuration. This diameter
matches the existing Contraves
-built 5.4m payload fairing,
thus eliminating many structural elements and permitting the
vehicle to fly from existing launch complexes with minimal
modifications. This design reverses the internal common bulkhead
and streamlines many systems while permitting many existing systems
and components to fly unchanged. Modular design enables multiple
engine configurations from one to six RL-10s or up to three
advanced next-generation high performance engines. This evolved
Centaur can be flown either on existing Atlas boosters ( designated
a Phase 1 configuration) or on next generation 5.4m diameter
boosters (designated Phase 2).
Performance levels for the Evolved Centaur based Phase 1 vehicles
envelope all Atlas V capabilities. In certain circumstances a
single Atlas booster vehicle with five solids and with an evolved
Centaur upper-stage can replace a three-booster core Atlas V-Heavy
(HLV). This has
obvious reliability and cost benefits. Phase 2 vehicles open the
door to a vastly higher performance capability. Up to 80 metric
tons can be lifted to low earth orbit on a Phase 2 HLV vehicle — a
substantial fraction of a Saturn V or Ares V vehicle. This
performance level, mandated only by NASA crewed exploration
missions, can be achieved using hardware identical to that used for
traditional commercial and USG missions thus allowing development
and support costs to be diluted by rate.
Studies have been conducted showing the extensibility of the basic
Centaur and Evolved Centaur designs to long duration space flight
for exploration purposes and even for use as a Lunar Lander.
Complementing these basic performance capabilities is the ability
to rate the vehicle for crewed operation. Extensive work has been
conducted showing that achieving this "man-rating" is
straightforward and does not mandate wholesale design changes to
the Centaur vehicle.
Test bed for cryogenic fluid management experiments
Lockheed Martin Space Systems has described the ability to use
existing Centaur hardware, with little modification, as a test bed
for in-space cryogenic fluid management techniques. Most Centaurs
launched on Atlas have excess propellants, ranging from hundreds to
thousands of pounds, which could be used for “rideshare”
experiments flown as secondary payloads conducted after separation
of the primary spacecraft.