In the late 1950s, American rocket engine technology was just leaving its juvenile years growing up rapidly with the arrival of the Red Scare. From its infancy to intensive study and derivatives gained from the V-2 missile and its scientists, American rockets began to play with fuels with greater energy for more efficiency as they concentrated on the United States Government’s central rocket objective: Intermediate-range and intercontinental ballistic missiles.

These engineers were brave, if not foolhardy people, experimenting with very exotic fuels and oxidizers. Some of them included liquid fluorine and (holy crap) chlorine trifluoride. Go ahead and look that one up. I’ll wait. Or you can enjoy this brief video that details how that chemical can make asbestos catch fire.

On one test firing using fluorine-based oxidizers, the engine melted.

Fluorine and derivative chemicals are highly reactive with basically anything, including glass. Mishandle it, and you get a very toxic and hot fire. Since basically anything reacts to fluorine (including metal, glass, water, sand and fire suppressants), your best firefighting plan if fluorine got loose, as rocket fuel developer and writer John D. Clark noted, was “a good pair of running shoes.”

While fluorine could provide a lot of energy, the risks clearly outweighed its benefits.

The reliable fuels for most rockets leaned to denser, less reactive chemicals, such as RP-1 (a type of kerosene), other non-fluorine hypergolics (chemicals that burn on contact) or sticking to what worked from von Braun’s A4/V2 design: liquid alcohol, combined with liquid oxygen as the oxidizer.

A few remained delighted to push the envelope, although not so far as going with insane hyper-hypergolic things like fluorine. They looked to a fundamental element, hydrogen. There was no reason for future rockets to forsake the immense power of that fuel. It was just a “simple” matter of design.

Convair_SM-65F_Atlas_532_550_SMS_Site_02_Abilene_KSBy 1959, the first cryogenically fueled American rocket engine was in its first test firings with Pratt & Whitney. The goal for this engine: Provide power as a second stage on a few rockets able to lift them–the Atlas and Titan rockets during the late 1950s to early 1960s.

While today’s rockets commonly have multiple stages (or secondary rocket boosters such as side-mounted solid rockets), virtually all of the earliest American rockets were single-stages since, before NACA and NASA’s involvement, these rockets were built as only suborbital, intercontinental ballistic missiles.

The original “stage and a half” Atlas itself had plenty of power to push itself into low earth orbit with sufficient speed to hurl a nuclear warhead. A true second stage was meant to do more with Atlas’ lifting power than pummel another country with its payload.

The Atlas provided the young NASA the greatest versatility because of the Air Force’s continual speedy expansion of the US missile stockpile, which included their plans to replace the Atlas over time with the Titan rockets. NASA would have quite a number of Atlas vehicles for any application, including the manned orbital phase of Project Mercury.

There were dreams of flight beyond earth orbit, of course. Space probes to Venus and certainly Mars. And there was the moon. No single-stage rocket built to that date could carry a payload into earth orbit with sufficient power to accelerate itself to escape earth gravity to reach cislunar and interplanetary destinations. A powerful second stage was required.

Agena S66-24482-orig
Atlas-Agena GATV-5003 lifts off for eventual rendezvous with Gemini 8. (NASA)

There was only challengers to the second stage throne. The most promising one used was the Agena. A great success in military applications, NASA would press Agena into use for a few lightweight space probes as well as converting the stage into a rendezvous target vehicle for use with Project Gemini. Delays in development of the new cryogenic engine and the stage it would power, the Centaur, fueled skepticism of liquid hydrogen’s viability. The delays also favored development of the Agena as a Saturn upper stage.

Agena’s success was partially due to the use of hypergolic fuels and oxidizers. The Agena could be stored fully fueled and didn’t require much pampering to keep it warm or cool, making it comparably easy to use.

But Agena’s total lifting power was barely 14,000 pounds/thrust with perhaps five minutes of burn time. While Agena would be enough to send some of the earliest, small probes to the moon (Ranger) and even Mars (Mariner), the stage would not be enough for heavier payloads.

The new cryogenic engine, designated the RL-10, and the upper stage concept, the Centaur, had quite a few doubters. Among them was Wernher von Braun. No stranger to the challenges of balancing risk over performance of liquid rocket fuels, von Braun felt time was best spent on denser, more manageable but less energetic fuels already in use. Perhaps von Braun also recalled the early challenges in Germany of getting hydrogen gas safely liquefied, not to mention combustion issues and extra precautions to avoid a gas explosion. Von Braun was also never fond of the pressurized-balloon tank concept of Atlas and felt Centaur’s similar design would manifest similar flaws.

Things were improving–slowly– in hydrogen’s favor. In the late 1950s the facilities and resources required to bring hydrogen and oxygen gases down to cryogenic temperatures were sufficient and safe enough to make bold steps. A failed Air Force project to develop a liquid-hydrogen powered aircraft also fed into developing the fuel infrastructure and hardware eventually adapted for the RL-10.

The first Atlas-Centaur test flight in May 1962  ended almost as quickly as it started, when an insulation panel on the side of the Centaur fell off. Its fuel overheated and the tank failed, raining shrapnel that pierced the Atlas, disintegrating the vehicle.

Inquiries on the usefulness of Centaur reached the presidential level, where NASA’s administrator, James Webb, managed to defend the project from cancellation by President Kennedy.

RL-10 461641main_gpn_original_full_full
Engineers test the RL-10 engine in NASA Lewis (now Glenn) Research Center’s Propulsion Systems Laboratory. (NASA)

NASA’s Centaur team, led by Lewis Research Center’s Abe Silverstein and others, kept at it, despite serious development and administrative gaffes that delayed the program, leaving several probes forced to reduce their sizes and weights to fly on Agena stages in Centaur’s absence.

With a successful second launch a few days after the assassination of President Kennedy in 1963, even skeptics like von Braun were eventually convinced of the possibility, if not the immediate practicality of liquid-hydrogen. Centaur and the RL-10 led to the decision to use liquid hydrogen fueled stages on all upper Saturn rockets, although von Braun never warmed up (or needed to consider) Centaur’s use atop any Saturn vehicle.

Plans to consider Centaur as an upper stage atop Saturn or even the lunar direct-ascent Nova concept disappeared as von Braun changed his mind on the best mode for landing men on the moon by 1962.

Development of the RL-10 wasn’t limited to Centaur. Just a few months later, Saturn’s first test of hydrogen used six RL-10 engines to form the Saturn S-IV stage used with the Saturn I launch vehicle, launched on the SA-5 mission on January 29, 1964. While elegant in its design and with up to three times the lifting power of the Centaur, the S-IV stage was used only a few flights, primarily as part of the Pegasus micro-meteoroid satellites. In its short lifetime, the S-IV helped further solidify the reliability of cryogenic power on future Saturn upper stages.

The Saturn I S-IV stage and the gorgeous RL-10 six engine cluster, generating 90,000 pounds/thrust. (NASA)

Centaur and the RL-10’s success record was not strong out of the gate, at only 80%. Some additional failures after Atlas-Centaur 2 managed problems with insulation, propellant sloshing and gas issues. A reaction control system was added to allow the stage to make a ullage burn prior to engine restarts as well as finer course control. With deadlines looming, Centaur was again on the chopping block before an Atlas-Centaur made a perfect test flight with a simulated payload. There were still significant problems with restart control of the RL-10 right into 1966, when an updated engine was introduced.

Centaur 660696main_1964_Centaur_hi-res_full
An early Centaur stage being lowered into a testing chamber. (NASA)

The seven Surveyor probes were Atlas-Centaur’s first operational missions and its greatest first triumph, hurling most of the lunar lander probes successfully towards the moon by direct ascent, without orbiting earth. While only five of the Surveyors made it to a soft landing on the lunar surface, what faults occurred were within the probes, not the launch vehicle. Surveyor sent back surface photos and soil data from the moon, verifying that manned missions could safely land without dropping into a sea of dust or shredded by crater walls and boulders. Centaur was fully operational, thanks to the Surveyor missions, although not fully mature.

The RL-10 engines were successfully restarted in space on the Atlas-Centaur 9 test mission. This meant coasting could be leveraged outside of America’s summer seasons to send spacecraft anywhere they desired, including the moon.

For a brief, shining period, the United States had a powerful family of engines (with increasingly diverse launch vehicle families) for any stage, most any fuel and any need, lifting things to low to high earth orbit or to the moon itself. As it turned out, RL-10 and the Centaur would long outlive all of its immediate successors.

Of these four engines, all used at some point during Apollo-Saturn development, only the first and smallest of these engines continues its work, over 65 years after it was developed. (NASA)

But as the Apollo program fulfilled its mission by the late 1960s, fewer orders for Centaur was ordered up. General Dynamics, manufacturers of the Centaur and Atlas, thought that their time was up, especially given the buzz about reusable rockets to save costs.

While those reusability ideas were being fleshed out for the post-Apollo era, NASA’s Jet Propulsion Laboratory had other ideas that wouldn’t wait for the proposed reusable launch vehicle. They wanted to explore the solar system unlike ever before–and only Centaur offered an opportunity to make that happen in time.

Thanks to the reliability and updates for the RL-10, the Centaur became the principal lifter for planetary probes. These new orders required significant modifications that, over time, stretched the Centaur’s length, improved its guidance computers and insulation to allow it to stay operating for several hours rather than minutes, increasing its versatility to sling the heaviest payloads almost anywhere the customers wanted.

The Pioneer 10 and 11 probes were launched by Atlas-Centaurs in early 1972 and 1973 while the Voyager 1 and 2 probes were launched with the more capable Titan III-Centaur in 1977, all as part of a rare alignment of the outer planets known popularly as the “Grand Tour“, allowing gravitational slingshots of one planet to reach others.

This video from the 1970 is a nostalgic summary at how well the Atlas-Centaur served as the central launch vehicle for most needs throughout the golden age of spaceflight, before work spread of an entirely new manned spacecraft meant to make all conventional rockets obsolete.

Atlas and Titan vehicles along with Centaur stages were expected to become obsolete with the arrival of the Space Transportation System–the Space Shuttle. Expected to be a far cheaper way to send anything into earth orbit and beyond, NASA succeeded in convincing the Air Force and Congress of the Shuttle’s potential, and production of conventional rockets looked ready to ramp down, with Centaur also poised to be junked or left as relics in museums.

But Centaur and its RL-10 engines could still be needed even in the Shuttle era. For all its promise, the Shuttle Orbiter lacked a high-energy stage to send its large payloads out of earth orbit.

Plans were made for a specialized Centaur to be flown inside the Shuttle’s cargo bay. In orbit, this Centaur would send several new interplanetary probes quickly on their way to Venus and Jupiter.

The loss of the Orbiter Challenger in 1986, after only 24 previous Shuttle flights, would soon rethink the safety of carrying the thin-walled Centaur in future Shuttles, and the idea was abandoned. Conventional launch vehicles, including a fully-redesigned Atlas and the ever-versatile Centaur, were back in production as further Shuttle commercial and military satellite launch plans were all but dropped. What few satellites launched by Shuttle after 1988 used a far less powerful two-stage solid rocket, which would add months to a space probe’s journey.

Centaur continued its hauling. The Cassini/Huygens Saturn probe left for the gas giant under the power of a Titan IV-Centaur in 1997. Latter Delta III and IV launch vehicles used a single RL-10 for their cryogenic upper stage. Together with the hypergolic Aerojet AJ-10 on earlier Delta vehicles, these stages launched many smaller probes over the years, including Pathfinder in 1996, the two Mars Exploration Rovers Spirit and Opportunity in 2003.

Centaur justified itself yet again as the end of the millennium came and went.

Centaur TDRS KSC00pp0704-orig
A Centaur with two RL-10 engines, in integration for a Tracking and Data Relay Satellite launch in 2000. (NASA)

Now riding on the Atlas V, a totally redesigned launch vehicle with little of its ballistic missile heritage, the Centaurs still are built today, with one or two RL-10 engines as needed, nearby their Atlas stages at the United Launch Alliance factory in Decatur, Alabama. While the Atlas V has dispensed with the always-pressurized balloon design of its predecessors, the Centaur retains this feature to maximize its payload lifting value.

The RL-10 engine itself has been upgraded over the years. The current engines used for the last Delta flights as well as current Atlas V missions generate 22,000 to 25,000 pounds/thrust, notably more than the original 15,000 pounds/thrust design. Combined with the Atlas V’s lifting power (often augmented by solid rockets), today’s Centaur is typically powered by a single RL-10.

While the engine’s total thrust hasn’t dramatically changed, the RL-10 makes the most of cryogenic fuel, able to restart and burn for longer periods to meet any needs asked of it.

TDRS-M Liftoff
Atlas V-Centaur launch of the TDRS-M satellite on August 18, 2017. I had the pleasure of seeing this launch vehicle under construction at the ULA factory, months before, in April. (NASA)
Centaur goesr_11
A telemetry-based rendering of the GOES-R Centaur and its single RL-10, relighting for a third time to send the satellite to its geostationary destination on November 19, 2016. (NASA/SpaceflightNow)

Centaur stages will even fly for a time with the next-generation Vulcan launch vehicle, scheduled to replace the Atlas sometime on or after 2019.

An RL-10 engine on display at the U.S. Space and Rocket Center in Huntsville, Alabama. (Personal photo)

However, after nearly 60 years of service, the venerable RL-10 and Centaur may finally be phased out with the Advanced Cryogenic Evolved Stage and its Blue Origin-designed BE-3 engine.

If there was ever an awards program in rocket engine history, the RL-10 would certainly earn the lifetime achievement award. No other engine has served the American space program in both civilian and military applications with greater accomplishments. It is believed that the Soviet Union’s early work in cryogenic engines may have flown with their N-1 moon rocket. But the N-1’s failures would delay the Russian’s entry into high-energy upper stages for decades, limiting their ability to reach beyond earth orbit.