Giving Mercury Its Wings: The First Test Flights of NASA’s Mercury Program

As 1959 opened, the newly created NASA appeared to be well on its way with Project Mercury. In January 1959 the Space Task Group (STG) based at NASA’s Langley Research Center in Virginia had chosen McDonnell as the capsule’s contractor and had already made key decisions on its design (see “The Origins of NASA’s Mercury Program”). While a modified Atlas ICBM would eventually launch Mercury into orbit, it was recognized early on that smaller rockets would also be required for various test flights. As soon as NASA was formed and Project Mercury organized in October of 1958, STG officials had begun negotiations to procure the rockets they needed.

 

The Redstone

One of the smaller new rockets required was the Chrysler-built Redstone. This short range tactical missile was developed by a team at the ABMA (Army Ballistic Missile Agency) under the direction of Wernher von Braun (see “Redstone: The Missile that Launched America into Space”). Since its inception in August 1953, the Redstone had been developed into a highly reliable rocket. A modified version of the Redstone was used by the von Braun and his team as the basis of a launch vehicle that sent America’s first satellite into orbit (see “Explorer 1: America’s First Satellite”).

A comparison of the Redstone missile, the Jupiter C and Mercury Redstone Launch Vehicle. Click on image to enlarge. (NASA/MSFC)

Among the many proposals to send men into space that circulated during 1958 was von Braun’s Project Adam. It proposed to send a man on a suborbital flight using a Redstone rocket. While the Project Adam proposal was rejected, the idea of using the Redstone to test Mercury in a short suborbital flight was attractive. A modified Redstone would be capable of reaching a speed of 2,000 meters per second and a peak altitude of about 185 kilometers – roughly orbital altitude. A total of 5 minutes of weightlessness would be experienced during the 15-minute flight.

ABMA’s “Man Very High” or Project Adam proposed by Wernher von Braun’s group would have launched an astronaut on a suborbital spaceflight using a Redstone missile. Click on image to enlarge. (ABMA)

The version of the Redstone to be used for the Mercury program, known as the Redstone MRLV (Mercury Redstone Launch Vehicle), started with the Redstone variant with stretched propellant tanks used on the Jupiter-C/Juno I rocket. But instead of using the highly corrosive and toxic Hydyne as a fuel (a mixture of 60% unsymmetrical dimethylhydrazine (UDMH) and 40% diethylenetriamine (DETA)), the Redstone MRLV reverted to the lower-performance (and safer) 75% ethanol mixture used in the tactical Redstone model. These propellants would feed a Rocketdyne A-7 engine (the latest version being used on production Redstone missiles) to generate 350 kilonewtons of thrust for about 145 seconds. In all, a total of 800 changes would be made to the Redstone to make it suitable for supporting manned flights.

Diagram of the Redstone MRLV. Click on image to enlarge. (NASA)

On January 16, 1959 STG placed an order with Chrysler for eight man-rated Redstone rockets. Two of the rockets would be used for unmanned test flights to qualify Mercury-Redstone combination, develop checkout procedures and gain experience. The remaining half dozen Redstones would be used for manned suborbital test flights allowing all six of the anticipated Mercury astronauts to each make a short foray into space (in the end seven astronauts were actually chosen but the Redstone order remained unchanged – see “Project Mercury: Choosing the Astronauts and their Machine”). The first Mercury-Redstone launch was planned for December of 1959 with the first manned attempt taking place the next month – only one short year away.

Cutaway diagram of the Jupiter IRBM developed by Wernher von Braun’s team at ABMA. Click on image to enlarge. (ABMA)

Along with the Redstone rockets, STG also ordered a pair of Jupiter IRBMs from ABMA. The larger Jupiter missile, which was designed to hurl a nuclear warhead over a range of 2,400 kilometers, would be capable of accelerating the Mercury capsule to a peak speed of 4,900 meters per second. While still short of the 7,800 meters per second required to attain orbit, it would provide a realistic reentry test for Mercury’s innovative ablative heat shield. A single test of the Mercury-Jupiter would take place sometime in the last quarter of 1959 before the first Mercury-Redstone flight. The second Jupiter would serve as a backup in case of a problem.

 

The Little Joe

While the Redstone and Jupiter missiles allowed important tests to be performed prior to manned orbital missions, these launch vehicles were ill suited and too expensive for certain types of tests. One of the more obscure of the originally proposed tests of the Mercury capsule required no launch vehicle at all. Instead a large balloon would carry a production version of the capsule to an altitude of 25 kilometers where it would be exposed to a near-space environment for extended periods. This would allow engineers to measure radiation levels and determine how the spacecraft operates under cold conditions.

Diagram showing the Mercury-Little Joe used for abort testing of the launch escape system. Click on image to enlarge. (NASA)

But other tests would require the use of specially designed rockets such as one known as “Little Joe”. This very simple yet adaptable launch vehicle was originally proposed by STG’s Max Faget for use in testing Mercury’s important tractor rocket launch escape system which he had developed. The fin stabilized Little Joe had no guidance system and used clusters of solid rocket motors to propel a Mercury capsule and its attached launch escape tower to speeds and altitudes that simulated critical flight conditions that would be experienced during an actual ascent. The escape system would then be activated and put through its paces. These flights would also allow the spacecraft’s recovery systems and procedures to be tested.

Diagram showing the internal configuration of rocket motors used in the Little Joe launch vehicle. Click on image to enlarge. (NASA)

Little Joe was designed to use various combinations of solid rocket motors to achieve different ascent profiles. The Little Joe was 15.3 meters tall and two meters in diameter with a fin span of 6.5 meters. Inside its airframe were mountings for four large rocket motors manufactured by Thiokol. Based on the Sergeant rocket, they came in two varieties: the XM-33E2 “Castor” (which produced a near-constant thrust of about 230 kilonewtons for 37 seconds) and the XM-33E4 “Pollux” (whose thrust increased from about 210 to 275 kilonewtons over a 24-second burn). Variants of the Castor would be used later as the second stage of the all-solid Scout rocket family and later as strap-on boosters for Thor-based launch vehicles. Four smaller Thiokol-built motors based on the Recruit rocket, designated XM-19E1-C12, supplemented the larger foursome at launch with a 1.6 second burn which produced an average of 160 kilonewtons of thrust. These eight motors could be ignited in various combinations at preprogrammed times to achieve a desired flight profile. In theory Little Joe could propel a Mercury capsule to a speed of 2,900 meters per second and reach heights in excess of 160 kilometers. Even though it could supply as much kinetic energy as a Redstone, Little Joe was not suitable for manned flights due to the lack of a guidance system and the inability to shut off solid motors once lit. Little Joe’s major advantage over the Redstone was price: Each Little Joe cost one-fifth as much as a Redstone.

Diagram showing a typical flight plan for the Mercury-Little Joe tests. Click on image to enlarge. (NASA)

Although it was capable of hurling a Mercury capsule into space, the Little Joe was used exclusively for tests of the important launch escape system at altitudes of 9 to 30 kilometers. Given this rocket’s low cost and flexibility, Little Joe was ideal for the task. In a typical test flight, two large and all four small solid rocket motors would ignite at liftoff. The remaining two large motors would ignite at predetermined times to achieve the goals of a particular test flight. Typical apogees for these flights, which would be launched over the Atlantic from NASA’s facility at Wallops Island, Virginia, were in the 15 to 85 kilometer range. On December 29, 1958 North American Aviation’s Missile Division got the contract from STG to build seven Little Joe rockets. At the beginning of 1959, the first Little Joe flight was set for July 1959.

 

The Atlas D

The Atlas, which would send Mercury into orbit, was a modified version of the operational Atlas D ICBM built by the Convair division of General Dynamics. The highly innovative Atlas used an integral balloon tank design where the millimeter-thick, stainless steel structure acted as both the outer shell and propellant tanks with internal pressure providing the rigidity needed to keep it from collapsing.

Diagram comparing the Mercury-Atlas and Atlas ICBM. Click on image to enlarge. (NASA)

Convair, working together with engineers from North American Aviation’s Rocketdyne Division (today part of Aerojet Rocketdyne), developed an innovative engine arrangement to power the Atlas. The MA-2 propulsion system used by the standard Atlas D ICBM consisted of a pair of Rocketdyne LR-89 boosters generating 687 kilonewtons of thrust each and a central LR-105 sustainer engine rated at 253 kilonewtons. All three of these engines would ignite on the launch pad to get the Atlas off the ground. After the ascending rocket had shed enough mass and gained sufficient altitude, it would jettison the pair of booster engines and their supporting structure. Greatly lightened, the Atlas would continue to accelerate towards its distant target powered by the single sustainer engine and the pair of verniers feeding off the remaining RP-1 grade kerosene and liquid oxygen (LOX) propellants in the tanks. This stage-and-a-half design along with the lightweight structure gave the Atlas D ICBM an impressive range of about 15,000 kilometers.

Diagram showing the configuration of the Rocketdyne’s propulsion system developed for the Atlas. Click on image to enlarge. (Rocketdyne)

The first Atlas A test flight was launched on June 11, 1957 with poor results (see “The First Atlas Test Flights”). As lessons were learned from the Atlas A, the Atlas B started it test flights on July 19, 1958. The launch of a stripped down “hotrod” version of the Atlas B into orbit on December 18, 1958 as part of Project SCORE demonstrated that the Atlas could orbit a payload (see “Vintage Micro: The Talking Atlas”). The first Atlas D test flight was launched (and failed) on April 14, 1959. The version of the Atlas D to be used for Mercury included a number of changes such as an abort detection system. Incorporating all the upgrades of the D-model plus new features required to man-rate the rocket, the Atlas-D would be capable of placing a 1,350 kilogram capsule into a 210-kilometer orbit.

The launch of Atlas 3D from LC-13 at Cape Canaveral for the first Atlas D test flight on April 14, 1959. (USAF)

Even though the Atlas D was critical to Project Mercury, it was also vital to national defense in its intended role as an ICBM. As a result, NASA projects including Mercury had to compete with the USAF for Atlas D rockets coming off Convair’s assembly line. NASA’s STG placed the first order for an Atlas from the Air Force Missile Division on November 24, 1958. A single, more readily available Atlas C was initially ordered to perform a suborbital reentry heating test using an ablative heat shield-equipped mockup of the Mercury capsule which would become known as “Big Joe”. With the availability of the Atlas D the following month, NASA modified its request ordering a total of nine D-models and deleting the interim C-model. Later more were added bringing the total order to 14 Atlas D missiles.

At the beginning of 1959, the Big Joe suborbital reentry test was scheduled for July of 1959. The first unmanned orbital test was scheduled for January 1960 and a manned orbital attempt that April. While in retrospect such an aggressive development schedule would have guaranteed that the US would launch a person into space before the Soviet Union, it proved to be far too ambitious.

 

Falling Behind

As 1959 progressed, STG officials were beginning to experience a degree of sticker shock with Mercury as costs began to spiral out of control. By the spring, McDonnell’s original $18.3 million estimate for a dozen capsules that NASA had ordered quickly skyrocketed to $41 million including spares and support equipment. Similarly, Mercury’s launch vehicles also began climbing in price. In January of 1959, the USAF raised the unit price of an Atlas by 32% to $3.3 million. Similarly, ABMA also increased the price of the Redstone and Jupiter. With a Jupiter now costing as much as an Atlas, the Mercury-Jupiter flight was finally scrapped on July 1, 1959. In a further move to control costs, the balloon flight was also canceled with a cheaper test to take place inside a modified wind tunnel.

With the full scope of the difficulty in developing the Mercury capsule and supporting hardware beginning to dawn on STG and NASA officials, the ambitious flight schedule also began to slip drastically. Even by March of 1959 NASA had slipped the first manned Mercury-Redstone flight originally planned for January 1960 to no earlier than the end of April. The first manned Mercury-Atlas flight was now pushed back to September of 1960. But within a month these dates had slipped again with the first manned Mercury-Atlas flight now not expected to take place before May of 1961 – a delay of 14 months just in the opening four months of 1959.

The first test flight of the Mercury program would use the Little Joe booster to assess Mercury’s launch escape and recovery systems. For these early test flights, inexpensive boilerplate models were employed instead of actual Mercury spacecraft. Boilerplate models mimic the mass, shape and dynamic properties of flight models but otherwise only carry systems and instruments needed for the tests being conducted. Their low costs and adaptability make them ideal for early testing of a new spacecraft design.

Mercury boilerplate models shown under construction in the workshops at NASA Langley Research Center. (NASA/LRC)

The Mercury program’s boilerplate models, whose development started even before McDonnell got its contract, were built in-house at Langley. For the flightworthy boilerplates to be used in the Little Joe and Big Joe test flights, most of the exterior was constructed at Langley from corrugated sheets of Iconel (a heat resistant nickel-chromium alloy) riveted together with interior stiffening rings. Emulating the size and shape of the actual spacecraft being built by McDonnell, the boilerplates were truncated cones with a base diameter of 1.9 meters and a height of about 2.3 meters. Inside was a fiberglass pressure vessel for instrumentation taking up about half of the interior volume and a parachute system in the cylindrical nose which reproduced the Mercury design. The base, which included a heat shield, along with the telemetry systems were constructed at NASA’s Lewis Research Center in Cleveland, Ohio (now the Glenn Research Center, named after Mercury astronaut John Glenn) and shipped to Langley for installation. The masses of these boilerplates were about one metric ton and varied from flight to flight depending on the equipment carried to support each mission.

Diagram showing the major components of Mercury’s launch escape system. Click on image to enlarge. (NASA)

The launch escape system to be tested in these flights was composed of current Mercury flight components. It consisted of a three-meter tall lattice structure made of stainless steel tubing topped with a model 1-KS-52000 solid rocket motor built by the Grand Central Rocket Co. (which was acquired by Lockheed in 1961). This 38-centimeter in diameter motor was 1.3 meters long and produced a nominal thrust of 231 kilonewtons for 0.78 seconds. Its exhaust passed through a trio of nozzles canted outwards by 19° to minimize its effect on the capsule. Beneath this was a small tower separation motor model 1.4-KS-785, manufactured by the Atlantic Research Corporation which produced 3.5 kilonewtons of thrust for about 1.3 seconds. The entire launch escape system, including a half-meter aerospike at the top, was 4.8 meters tall with a total mass of 460 kilograms. The launch escape system was first tested successfully on July 22, 1959 in a beach abort test at Wallops. The 984-kilogram boilerplate capsule hit a peak altitude of 594 meters before successfully deploying its chutes for a safe splashdown offshore.

A boilerplate capsule with the launch escape tower attached being prepared for an abort test mission. (NASA/LRC)

The first Little Joe flight, designated LJ-1, was originally planned for launch in July 1959 but it was not until August 21 that everything was finally ready. The objective for this flight was to test the launch escape system during an abort when the ascending spacecraft was experiencing maximum dynamic pressure or “max-q”. This would be the most stressing scenario expected during the ascent into space. This Little Joe used four Pollux motors and a quartet of Recruit auxiliary motors. With LJ-1 on its pad tilted towards the Atlantic, the countdown proceeded until the T-35 minute mark when the launch area was evacuated in preparation for launch.

Crews preparing the Mercury launch escape system and boilerplate model for stacking atop of the Little Joe rocket on August 18, 1959 for the LJ-1 mission. (NASA/LRC)

Unexpectedly, the launch escape system fired about 31 minutes before the scheduled launch and pulled the boilerplate Mercury from its booster and out to sea in an unintended pad abort test. The clamp holding the escape tower to the capsule released near apogee at an altitude of 600 meters as was safely pulled away by the separation motor. While the 1.8-meter drogue chute deployed during descent of the capsule, there was insufficient electrical power available to release the 19.5-meter main parachute. The boilerplate capsule crashed into the sea 800 meters downrange after a flight of only 20 seconds.

The Little Joe rocket left behind after the accidental activation of the launch escape system during the countdown for LJ-1. (NASA/LRC)

The subsequent investigation showed that an electrical transient had triggered a relay specifically designed to prevent too rapid an abort as well as the destruction of the Little Joe rocket. In the end, the LJ-1 mission was judged to be too ambitious. A new test, designated LJ-6, was inserted into the test plan. Using the undamaged Little Joe rocket from the LJ-1 test, the LJ-6 mission would concentrate on testing the Little Joe booster itself. Afterwards, the max q abort test would be attempted once again.

 

The Big Joe Flight

With this inauspicious start to Mercury’s flight test program, engineers continued their preparations for the Big Joe Mercury-Atlas ballistic test flight. The purpose of this flight was to test the new ablative heat shield which had been selected for use on orbital Mercury flights. Secondary objectives included verifying the aerodynamics of the Mercury shape in flight as well as gaining experience with all aspects of a Mercury-Atlas flight from launch to recovery. Langley built a pair of specially prepared boilerplate capsules based on the design used for the Little Joe flights for Big Joe each with a mass of 1,159 kilograms. If the first Big Joe flight failed to meet its objectives, the second boilerplate would be launched.

These drawings from February 1959 showing the Little Joe boilerplate design (right) and the originally proposed Big Joe boilerplate design (left) and the design actually flown. Click on image to enlarge. (NASA)

Engineers at Lewis built a cold gas attitude control system for the capsule which used about a kilogram of pressurized nitrogen (instead of the “hot gas” hydrogen peroxide-based system to be used on Mercury and employed on the X-15 – see “The First Reusable Spacecraft: The Origins and First Test Flights of the X-15”). The jets used for yaw and pitch control produced about 45 newtons of thrust while the roll jets produced about nine newtons each. The boilerplate capsule was fitted with 139 temperature sensors to monitor the conditions at key points on the exterior and interior during reentry. Of these, 52 were mounted on various points of the 2.7-centimeter thick ablative heat shield itself. Another 50 instruments recorded sound levels, pressures and the acceleration. All these sensors transmitted their findings back to Earth using a telemetry system developed by the engineers at Lewis.

Photo showing the major features of the Big Joe boilerplate capsule. (NASA/LRC)

The plan was for the mission’s launch vehicle, Atlas 10D, to liftoff from LC-14 at Cape Canaveral, Florida and fly down the Atlantic Missile Range. The Atlas would boost its payload (which would not include a launch escape tower) to an altitude of almost 160 kilometers before pitching downward to drive the spacecraft back towards the Earth. Atlas’ sustainer engine would shut down at an altitude of 140.0 kilometers and the boilerplate would be released at a speed of 7,282 meters per second on a path directed 1.50° below the local horizontal. This trajectory was designed to simulate a shallow reentry from orbit which would maximize the total heat load on the spacecraft. Once released, the capsule would use its attitude control system to turn itself around bluntside first for reentry and maintain a slow roll rate of one rpm. Since the Atlas would place the boilerplate capsule into the proper descent trajectory for this test, no retrorocket package was required.

A schematic of the Big Joe flight plan. Click on image to enlarge. (NASA)

Following reentry, the capsule would automatically deploy its drogue chute at an altitude of about 14 kilometers followed by the main parachute at three kilometers. Splashdown was expected to occur 3,300 kilometers downrange after a total flight time of 20 minutes. The spacecraft would then be recovered by one of a flotilla of a half dozen US Navy destroyers strung out along the expected ground track and brought back to Cape Canaveral to inspect the condition of the boilerplate and its innovative heat shield. The Big Joe mission would be the first full scale flight operation of the Mercury program.

Big Joe on the pad at LC-14 being prepared for launch. (USAF)

During the second week of June, Langley’s boilerplate capsule for the Big Joe mission arrived at Hanger S at Cape Canaveral sharing space with the team preparing the final two Vanguard rockets for launch (see “Vintage Micro: The Original Standardized Microsatellite“). While some hoped for the Big Joe liftoff on July 4, the launch was pushed out to mid-August by the USAF because of issues with Atlas 10D. NASA engineers then pushed the launch out further to early September to attend to problems with the complex instrument and telemetry systems in their capsule. By the evening of September 8, Atlas 10D with its boilerplate Mercury capsule stood ready for launch scheduled for 3:00 AM EST on September 9, 1959. At 2:30 AM EST, a 19-minute hold was called to investigate an issue with the ground guidance computer. Finally, at 3:19 AM, Atlas 10D lifted off lighting up the nighttime sky. This was just the sixth launch of the still new Atlas D.

The launch of the Mercury Big Joe mission from LC-14 during the early-morning hours of September 9, 1959. (USAF)

All seemed to be going as planned until about two minutes into the flight. After the booster engines shutdown as expected, they failed to jettison. Carrying the extra mass of its booster section, the Atlas struggled to keep itself on course but to no avail. When the sustainer finally shutdown upon the depletion of its propellant 293 seconds after launch, the spacecraft was at an altitude of 149.7 kilometers travelling in a path angled 0.92° below the local horizontal at a speed of only 6,287 meters per second – 994 meters per second slower than planned. To compound the velocity shortfall issue, the release of the boilerplate capsule from its spent Atlas was delayed by 138 seconds for some reason. The tiny capsule’s attitude control propellant supply was quickly exhausted as it tried in vain to pitch the capsule along with the dead weight of the Atlas around for reentry during those two minutes.

Without any means of reorienting itself, the descending capsule started its reentry after its delayed release. Despite the string of failures which threatened the mission, the robustness of the Mercury design was ultimately proven. Even without an active attitude control, the capsule righted itself during reentry owing to its shape and the careful placement of its center of gravity. While the slower reentry velocity lowered the total heat load the capsule experienced, the trajectory brought the descending craft into the lower atmosphere more quickly resulting in a higher peak heating rate. The Big Joe capsule survived reentry and splashed down in the Atlantic after a flight of 13 minutes some 2,407 kilometers downrange – about 800 kilometers short of its planned landing zone.

The boilerplate capsule from the Big Joe mission on the deck of the USS Strong following its recovery. (USN)

At local sunrise, the Big Joe capsule was spotted by a Navy P2V Neptune patrol plane allowing it to vector the nearest destroyer, the USS Strong over 160 kilometers away, towards the wayward capsule for recovery. By about 10:00 AM EST, the Big Joe capsule was recovered and brought aboard the USS Strong. It was subsequently transferred to a cargo plane and flown to Cape Canaveral. The capsule arrived by 10 PM on launch day and was moved to Hangar S by midnight. Close inspection of the capsule showed that only 30% of the heat shield had ablated away and that the capsule was perfectly protected despite the punishing flight. While there was some buckling of the outer skin of the capsule where heating was especially intense and a couple of recovery hooks on Mercury’s cylindrical forward section were heavily eroded, there were few signs of thermal damage to the capsule.

Despite the problems encountered during the Big Joe flight, the new ablative heat shield had proven itself as did the aerodynamics of the Mercury capsule shape. As a result, a “Big Joe 2” mission was not needed and its Atlas 20D launch vehicle was released for use in NASA’s Atlas-Able program (see “NASA’s Forgotten Lunar Program”). With this (partial) success under their belts, the Mercury program team pushed forward with their test schedule.

 

Follow Drew Ex Machina on Facebook.

 

Related Videos

Here is some footage of the Mercury Big Joe launch on September 9, 1959:

 

 

Related Reading

“Project Mercury: Choosing the Astronauts and their Machine”, Drew Ex Machina, April 9, 2019 [Post]

“The Origins of NASA’s Mercury Program”, Drew Ex Machina, December 17, 2018 [Post]

 

General References

David Baker, The History of Manned Spaceflight, Crown Publishers, 1981

William S. Blanchard, Jr. and James L. Raper, Full-Scale Flight Test from Sea Level of an Abort-Escape System for a Project Mercury Capsule, NASA Technical Memorandum X-422, October 1960

Ronald Kolenkiewicz and John C. O’Loughlin, Performance Characteristics of the Little Joe Launch Vehicles, NASA Technical Memorandum X-561, September 1962

Emily W. Stephens, Afterbody Heating Data Obtained from on Atlas-Boosted Mercury Configuration in a Free Body Reentry, NASA Technical Memorandum X-493, August 1961

Loyd S. Swenson Jr., James M. Grimwood and Charles C. Alexander, This New Ocean: A History of Project Mercury, NASA SP-4201, 1966

Project Mercury – Preliminary Flight Test Results of the “Big Joe” Mercury R and D Capsule, NASA Project Mercury Working Paper No. 107, October 12, 1959

Project Mercury: Man-in-Space Program of the National Aeronautics and Space Administration, Report of the Committee on Aeronautical and Space Sciences United States Senate, December 1, 1959