Before World War II, the field of rocket technology development was dominated by small groups of enthusiastic visionaries such as the American Rocket Society as well as brilliant individuals like Dr. Robert Goddard who launched the first liquid propellant rocket from Auburn, Massachusetts in 1926 (see “Goddard’s First Rocket Patents – July 1914”). While the efforts of these and other groups worldwide were relatively modest, all that changed with the start of World War II when governments dedicated major resources to develop rockets into effective weapons of war.
The German A-4 rocket (better known as the V-2) was developed by a team led by Wernher von Braun. (NASA/MSFC)
The ultimate product of these wartime efforts, in terms of size and range, was the A-4 rocket (better known as the V-2) developed by a team led by aerospace pioneer, Werhner von Braun. Designed to hurl a one-ton payload of high explosives over a range of 320 kilometers, the 12.5 metric ton V-2 was by far the largest rocket ever developed up until this time and the first capable of flying into space. Following the surrender of Germany on May 8, 1945, the western Allies and the Soviet Union scrambled to secure advanced German technology and the people who developed it to bolster their own military capabilities. During one of these efforts, known as Operation Paperclip, the US Army was able to get 300 railcar loads of V-2 rockets and components as well as von Braun and many key members of his team and bring them back to the US. While the immediate goal of these efforts was for weapon development, the launch of “surplus” V-2 rockets from American soil starting in April 1946 from the White Sands Proving Grounds in New Mexico also opened the opportunity for scientific investigation of the near-space environment for the first time.
The Genesis of the Viking Rocket
In order to make the best use of the V-2 rockets for scientific research for what became known as the Upper Atmosphere Research Program, a joint US Army-USAF-US Navy committee (which included participation by various laboratories and universities) was established early in 1946 with Dr. Ernst Krause as its head. Dr. Krause, who directed the Rocket Sonde Research Branch of the Naval Research Laboratory (NRL) formed at the end of 1945, was a physicist who had worked on guided missile research for the US Navy during the war. While the V-2 was capable of hurling a metric-ton payload on a suborbital trajectory to altitudes of up to 160 kilometers and more, it was soon recognized that it was far from ideal for scientific research.
A surplus V-2 being prepared for launch in 1946 at the White Sands Proving Grounds in New Mexico. (USN)
Although the payload capacity of the V-2 was impressive, it was far larger than was needed for the scientific payloads then available. In order for the V-2 to maintain its stability during ascent, up to a half a ton of lead ballast needed to be carried. Another problem was the stability of the rocket after powered flight. While the V-2 used fins and graphite vanes in the exhaust plume of its main engine to steer itself, it tended to roll and tumble uncontrollably once the engine shutdown and the rocket moved above the sensable atmosphere where its fins were rendered useless. This limited the amount of useful data which could be returned by some instruments especially those requiring stable pointing like cameras and sensors used to study the Sun. Finally, there were a finite number of rockets available and the supply would be exhausted within a few years.
Anticipating the need to develop their own sounding rockets, Krause and his team at NRL’s Rocket Sonde Research Branch came up with two sounding rocket concepts in early 1946 to investigate issues of particular importance to the US Navy. One of these areas of interest was observing weather phenomena from high altitude which could affect naval operations. Of greater importance was studying the properties of the ionosphere (layers of ionized gas in Earth’s upper atmosphere) and how they were affected by outside influences. During this era before the introduction of communication satellites, long distance, over-the-horizon communications was made possible by short wave radio transmissions repeatedly reflecting off of the ionosphere above 85 kilometers altitude and the solid Earth to reach distant locations. Unfortunately, the properties of the ionosphere changed not only over the course of the day but without warning resulting in “fade outs” which limited the range of transmissions at times. While the Sun played a role in how the properties of the ionosphere changed, the exact mechanisms at work were still a matter of debate in the scientific community.
Cutaway diagram showing the major components of the Aerobee sounding rocket. Click on image to enlarge. (NASM)
The first of the sounding rocket concepts examined by NRL led to the development of the comparatively small, unguided rocket eventually known as “Aerobee”. This inexpensive, 540-kilogram rocket would be capable of sending a 60-kilogram payload to an altitude of 125 kilometers and would be suitable for many of the Navy’s investigations of the threshold of space including direct observations of the ionosphere’s “D layer” (which was present only during daytime at an altitude of 60 to 90 kilometers and attenuated medium and high-frequency radio transmissions) and its “E layer” (which persisted during all times of the day and extended from about 90 to 150 kilometers).
Milton Rosen was the head of NRL’s Viking rocket program. (NASM)
The second sounding rocket concept was the brainchild of NRL’s Milton Rosen. Rosen, who worked at NRL under Krause during World War II developing missile guidance systems, was an electrical engineer by training who took a keen interest in rocketry as information about the V-2 and other German missiles became available after the war. He envisioned the development of a much larger, guided sounding rocket, initially known by the name “Neptune”, with a design goal of sending a 500-pound (230-kilogram) payload on a suborbital path to an altitude of 500,000 feet (150 kilometers – well above the 100-kilometer Karman line marking the threshold of “space”) and smaller payloads to still higher altitudes allowing direct sampling of the ionosphere’s “F layer” (which extended above 150 kilometers altitude and made long distance short wave communications possible). Incorporating the latest advances in American rocketry as well as some new lessons learned from the V-2, Rosen’s Neptune rocket (which would be renamed “Viking” in 1947 to avoid a conflict with the Lockheed P2V “Neptune” maritime patrol aircraft) would be about a third of the mass of the V-2. Still, it would be the largest rocket ever built in the US up until that time and provide a far superior platform for high-altitude scientific investigations at much lower cost – an estimated $40,000 for the new rocket versus $175,000 for a V-2 launch (about $465,000 versus $2 million, in today’s money).
A 1947 illustration of altitude-payload performance and costs of various rockets. The “Neptune” (second from left) would be renamed Viking later that year. (NRL)
Because of Rosen’s enthusiasm for his large sounding rocket proposal, Krause put him in charge of the Viking program. But with his formal training limited to electronics, Rosen felt he needed a more in depth education in all aspects of rocket technology. As work ramped up in 1946 defining the requirements and selecting contractors for Viking, Rosen visited the California Institute of Technology (better known as Caltech) which ran the Jet Propulsion Laboratory (JPL) in Pasadena, California – the foremost American institution in rocket development at the time. Following his initial visit to JPL in March of 1946, Rosen took up a temporary teaching position at Caltech from September 1946 to April 1947 which allowed him to learn everything he could about all aspects of rocket technology development and testing at JPL.
The Viking Design
Before Rosen departed for JPL, Krause and his team drew up detailed specifications for Viking and sent them to five companies. NRL subsequently received bids from three of these companies – General Electric (GE), Douglas Aircraft and the Glenn L. Martin Company. Despite its lack of rocket experience, Martin’s bid for an initial batch of ten rockets was chosen on August 21, 1946 for two reasons. First, it was located in Baltimore, Maryland not far from NRL in Washington, DC allowing Rosen and his team to make frequent visits and work closely with the prime contractor as the Viking was developed and built. Second, the Martin team was composed of a small group (never numbering more than 50) of young and enthusiastic engineers who were eager to incorporate the latest technologies into this innovative rocket.
Drawing of Glenn Martin’s proposal for what would become known as the “Viking” rocket. Click on image to enlarge. (Martin)
Unlike the V-2 with its steel structure, Viking used lightweight aluminum throughout to improve its mass ratio (i.e. the ratio of its fully fueled to empty masses). Since the widest sheets of aluminum available at that time were limited to 100 inches (254 centimeters), the diameter of the initial version of the Viking was just 81 centimeters so that only a single weld would be needed to form the cylindrical body of the rocket. In order to fit the required load of about 3.6 metric tons of propellant into the rocket along with its engine, other consumables, etc., its total length (minus the payload) was about 12 meters. As a result, the Viking had a distinctively slender, pencil-like appearance. In order to further lighten the rocket’s structure, the Viking incorporated an integral fuel tank feature where the outer skin of the rocket also served as the tank itself. The oxidizer tank, which would be filled with cryogenic liquid oxygen (LOX), was mounted inside the framework of the rocket’s outer structure so that it could be insulated to limit boil off of the LOX and the buildup of ice on the exterior prior to launch.
Cutaway diagram showing the major components of the NRL’s Viking sounding rocket as built. Click on image to enlarge. (Martin)
While the Viking used fins to help stabilize the rocket during its initial climb out much like the V-2, it incorporated a new gimbal engine mount to steer the rocket instead of graphite vanes used by the V-2. The graphite vanes in the V-2 exhaust plume tended to disintegrate during powered flight leading to failures. The gimbal mounted engine under the control of a gyroscopic inertial guidance system avoided this failure mode and also resulted in improved engine performance. Viking would be the first large rocket to use a fully gimballed engine for steering. After engine burnout, the attitude of the Viking rocket would be controlled by cold gas jets feeding from a 24-kilogram supply of high pressure nitrogen so that its instrument payload could be kept oriented in the proper direction for its measurements – another major improvement over the V-2. Depending on the payload and its requirements, the total length of the Viking was 12.8 to 14.6 meters with a total launch mass typically in the 4.4 to 5.2 metric ton range.
Diagram showing the operation of a rocket engine gimbal. (Rosen)
The innovative Viking rocket would need an equally innovative power plant to produce at least 89 kilonewtons of thrust for a nominal burn time of around 75 seconds. In August 1946, the NRL team chose Reaction Motors, Inc. as the contractor for its engine which started work the next month. Founded by four members of the American Rocket Society in December of 1941, Reaction Motors had concentrated on the development of JATO (Jet Assisted Take Off) technology during World War II to help get aircraft airborne more quickly than they otherwise could with their own engines alone. Following the war, Reaction Motors began development of their Model 6000-series liquid propellant rocket engine (better known by the USAF designation, XLR-11). With a nominal thrust on the order of about 27 kilonewtons (depending on the exact model), the four-chamber XLR-11 was used in some early experimental rockets as well as in rocket-powered aircraft like the Bell X-1 which broke the sound barrier for the first time on October 14, 1947. NRL chose Reaction Motors in part because of their enthusiasm but also because they were located fairly close by in New Jersey.
This is Reaction Motor’s XLR-10 engine and its turbopump. The inset image shows a test firing of the engine in New Jersey. (NASA)
The power plant developed for the Viking, designated XLR-10-RM-2, was a regeneratively cooled engine where the fuel is circulated through tubes surrounding the whole engine to help lower the engine temperature before being injected into the combustion chamber. Although initially made entirely of steel, Reaction Motors engineers eventually substituted nickel for the interior lining because of its superior thermal and mechanical properties. The propellants for the engine were a 95:5 ethanol/water mix (compared to the 75:25 mix used by the V-2 and Reaction Motor’s XLR-11 engines) and LOX. These propellants were delivered to the XLR-10 combustion chamber at a rate of about 50 kilograms per second by means of a turbopump with a shaft power rated at 224 kilowatts at 10,000 rpm. A supply of about 127 kilograms of 90% hydrogen peroxide was fed over a catalyst bed at a rate of 1.38 kilograms per second during the course of the engine burn to produce high pressure steam to drive the turbopump. The turbopump exhaust was tapped to feed jets on Viking’s fins to control the rocket’s roll during powered flight while engine gimballing controlled pitch and yaw. The XLR-10-RM-2 produced about 92.5 kilonewtons of thrust at sea level and 110.5 kilonewtons in a vacuum. Although it generated about one third of the thrust of the V-2 engine with slightly lower efficiency, the XLR-10 was still the largest rocket engine ever built in the US up until this time.
With its many innovations to enhance performance, Viking (which received the designation RTV-N-12) was capable of launching a 900-kilogram payload to an altitude of about 135 kilometers while a 45-kilogram payload could be hurled as high as 385 kilometers. This would give NRL the capability to send useful payloads to higher altitudes than the V-2 could reach with a much more science-friendly rocket.
Preparing for the First Launch
By the time Rosen returned to NRL in the spring of 1947, work was already well underway developing the Viking rocket and its engine. Rosen worked closely with the teams at Martin and Reaction Motors trying to incorporate the lessons he had learned at JPL. One of his early challenges was instituting a new testing philosophy where not only individual components are checked before assembly, but the whole rocket would be check once assembly had been completed. This included a requirement that the complete Viking be loaded with propellants and static fired as part of an end-to-end countdown rehearsal. Although there was some resistance, eventually this and many other engineering practices were introduced to improve the chances of success.
As would be expected with a project like Viking, the schedule for the delivery of Viking 1 was delayed for months as engineers struggled to resolve problems with the innovative rocket. For example, Reaction Motors had originally planned for a 14-month development effort for the XLR-10 engine but it was not until October 1947 (almost 14 months after their bid was selected) that engineers began static testing of their first of an eventual three engine prototypes at a test stand built by the US Navy at the Picatinny Arsenal in Morris County, New Jersey. The first XLR-10 flight engine did not complete its acceptance test until September 21, 1948 when it was fired for 66 seconds at a thrust of 93 kilonewtons.
A failure of one of the early prototypes of the XLR-10 engine during a test firing in New Jersey. (Reaction Motors)
Martin was also running behind schedule as their staff struggled with their own problems. Dynamic testing of Viking’s all-important steering system began in the spring of 1948. By December (a full 11 months behind schedule), the fully assembled Viking 1 rocket was undergoing its last tests. Rosen wanted a test of Viking’s gimbal system with the engine firing before accepting delivery of the first rocket but, such tests would have added further delays. While Viking 2 already on the test stand would be subjected to this testing just after New Year’s 1949, Viking 1 would forego them. With an agreement that some minor incomplete items on the rocket would be addressed at the launch site, Viking 1 was shipped by rail from Martin’s factory on January 11, 1949.
Like the V-2, Viking would be launched from Launch Complex 33 (LC-33) at the White Sands Proving Grounds in southern New Mexico. The US Army had started construction at LC-33 in July 1945 eventually building a heavily-built concrete firing pit to deflect the rocket exhaust during launch, blockhouse for the launch crew, a moveable service gantry, observation towers and tracking facilities to receive telemetry from the rockets and their payloads. Viking would be launched from Pad ALA1 which included a launch pedestal over the firing pit and a moveable gantry to allow the Viking to be serviced before launch. Equipment to support Viking launches was already installed as were Navy-built facilities to support the ground crews. The rocket would not stand vertically on the pedestal above the firing pit but would be tilted by up to three degrees towards the north so that the rocket would travel downrange. If the rocket deviated too much from its prescribe trajectory during powered flight, the engine could be shutdown remotely by ground command – another feature the early V-2 launches did not have.
The blockhouse at White Sands’ Launch Complex 33 (LC-33). Visible in the background is a Viking rocket on its pedestal over the firing pit and the gantry pulled back to a safe distance. (Martin)
Viking 1 completed its 3,200-kilometer cross-country trip on January 17, 1949 when it arrived at the railhead at Oro Grande, New Mexico. The first week of February was set aside for rocket operations rehearsals with Viking 1 being bolted to its launch pedestal over the firing pit on February 1. After completing propellant loading tests and several simulated firings, Viking 1 was returned to the hanger on February 5 for an additional three weeks of testing and preparation.
A Viking rocket on its transportation cart being moved to a hanger. (Martin)
Viking 1 was finally erected on its firing stand on February 28, 1949 in anticipation of its first static test firing scheduled for March 7. Preparations for the static test began at 2:00 AM MST with a firing planned for 10:00 AM. Problems with the disconnect plug (which attached the rocket to ground equipment during the countdown) caused numerous delays with the static test finally scrubbed at 5:30 PM. The attempt on March 8 was also scrubbed due to continuing problems with the disconnect plug, hydrogen peroxide servicing and issues locating sufficient supplies of nitrogen gas pressurant. The 30-second static firing finally took place on March 11 but, a subsequent inspection of the tail compartment of the rocket showed that a fire had broken out there charring some pipe insulation. Apparently, the seal between the two halves of the turbopump had opened up allowing peroxide-rich steam to leak out. After attempts up until March 18 to fix the problem on the pad failed, the rocket was returned to the hanger.
A static test firing of the Viking 1 rocket at LC-33 in White Sands. The rocket exhaust is deflected towards the right by a concrete firing pit dug into the desert floor.
The now repaired Viking 1 was returned to the pad on April 22, 1949 for a test firing on April 25. After some delays, the second static firing was started but the engine was shutdown after only 20 seconds when some smoke was spotted. When a post-firing inspection showed no problems and with good data from the test firing, it was decided to push towards an actual launch scheduled for April 28.
The First Test Flights
With Viking 1 finishing its final checks, its 210-kilogram payload was attached to the rocket. It included an instrument suite to measure temperatures and pressures in the atmosphere as well as a camera. If the camera was found intact after “landing” (despite no provisions to recover the rocket or payload), the photographs would provide high altitude views of the Earth as well as measurements of the rocket’s attitude during coasting flight. The first Viking launch attempt on April 28, 1949 was postponed first due to weather and then by the truck carrying the LOX supply arriving two hours late. After numerous launch attempts starting at 11:44 AM MDT, the launch was scrubbed by problems with the LOX tank vent valves. With the countdown for the second launch attempt starting at 2:00 AM MDT on May 2, the launch was again scrubbed when the yaw servo attachment for the engine gimbal was found to be broken.
The Viking 1 rocket shown with its gantry prior to launch. (Martin)
The last launch attempt for May 3 started with pressure growing on the Viking team to clear LC-33 for other scheduled rocket launches. Problems with the LOX relief valves delayed the countdown at 8:30 AM then again at 9:10 AM. With pressure growing to get off the pad before clouds were predicted to close in at 10 AM and the LOX supply running lower than desired, the 4,377-kilogram Viking 1 finally lifted off at 9:14 AM MDT. While Rosen still hoped to achieve an altitude of at least 160 kilometers despite the low LOX supply, the XLR-10 engine shutdown early 54.5 seconds after launch with the rocket travelling at 1,052 meters per second. Viking 1 reached its apogee of 80 kilometers after 164 seconds of flight. The first Viking not unexpectedly broke up during reentry after 291 seconds of flight and showered a ten square kilometer area with debris.
The launch of Viking 1 on May 3, 1949. (USN)
Viking 2 arrived at White Sands in July 1949 as engineers puzzled over the cause of the early engine shutdown on Viking 1. During an inspection of Viking 2 following its 30-second test firing a month after arrival, signs of another fire were found just as had happened with the first static test of Viking 1. A turbine leak now became the major suspect for the Viking 1 problem. The first launch attempt for Viking 2 came at 11:29 MDT on August 26 carrying another 187-kilogram payload to measure the properties of the upper atmosphere. Unfortunately, a LOX leak in the turbine had frozen the peroxide valves shut preventing liftoff. With the problem fixed, the 4,529-kilogram Viking 2 lifted off from LC-33 at 9:57 AM MDT on September 6. Once again, the XLR-10 engine shutdown early after only 49.5 seconds of flight. Travelling only 815 meters per second at the premature cutoff of the engine, Viking 2 reached a peak altitude of just 51 kilometers about 133 seconds after launch. Viking 2 impacted the desert after a total flight time of 394 seconds.
The launch of Viking 2 on September 6, 1949. (NRL)
It was quickly determined that the seal between halves of the turbopump was failing as a result of the high operating temperatures and the vibration of launch. Reaction Motors fixed the problem by welding the two halves of the turbine housing instead of bolting them together. With this and other problems encountered during the first pair of flights fixed, Viking 3 arrived at White Sands in early January 1950 for the first launch attempt scheduled for February 7. Among the improvements made, Viking 3 now sported a new integral LOX tank design like that which was already being used for the fuel tank. It was found that lightening the rocket was more important than providing insulation for the cryogenic oxidizer. The change also resulted in a slightly larger LOX load with the ethanol tank slightly lengthened to hold more fuel. The lighter structure and slightly increased propellant load would help to improve rocket’s performance.
The improved Viking 3 shown being prepared for launch from LC-33 at the White Sands Proving Grounds. (NRL)
The first static test firing was delayed by a peroxide valve issue and was then cut short after only 15 seconds. Telemetry showed that while the engine was working well, problems with the control system were shaking the rocket. A second static test firing on February 6, 1950 was finally successful with a burn time of 68 seconds. With the static test completed, Viking 3 was prepared for launch with its 240-kilogram payload which included a solar spectrograph, Geiger counters to measure cosmic rays, instruments to study the temperature and pressure of the upper atmosphere as well as a camera to record the Earth and the attitude of the rocket. The nose of Viking 3 would separate during descent upon ground command to maximize the chances that the solar spectrograph (and its data recorded on film) other parts of the science payload could be recovered.
The launch of Viking 3 on February 9, 1950. (Martin)
Viking 3, with a launch mass of 5,012 kilograms, lifted off from LC-33 at White Sands at 2:44 PM MST on February 9, 1950. Unfortunately, the trajectory of Viking 3 drifted westward forcing ground controllers to shut down the engine early by remote command after 59.6 seconds of flight when the speed was 1,049 meters per second. Viking 3 reached a peak altitude of 80 kilometers some 169 seconds after launch. During descent, explosives were fired to separate the nose from the rocket. The rocket impacted the desert 420 seconds after launch. The mangled nose of the rocket was found and the photographic film from its instruments were recovered.
Although a myriad of problems had been encountered during the test firings and flights of the first three Viking rockets, its innovative design features like the fully gimballed engine were proving themselves. While more work was needed to get Viking to meet its potential, Viking 4 would get its chance to prove itself to be a useful scientific tool as well as a technology demonstrator with a planned launch from the deck of the US Navy’s USS Norton Sound in the Pacific Ocean planned for May 1950.
Follow Drew Ex Machina on Facebook.
Related Video
Here is a period newsreel about the launch of the Viking 2 rocket in September 1949.
Related Reading
“Goddard’s First Rocket Patents – July 1914”, Drew Ex Machina, May 15, 2014 [Post]
“Redstone: The Missile That Launched America Into Space”, Drew Ex Machina, April 26, 2016 [Post]
“Vanguard TV-3: America’s First Satellite Launch Attempt”, Drew Ex Machina, December 6, 2017 [Post]
General References
David H. DeVorkin, Science with a Vengeance: How the Military Created the US Space Sciences After World War II, Springer-Verlag, 1992
J.D. Hunley, US Space-Launch Vehicle Technology – Viking to Space Shuttle, University Press of Florida, 2008
E.H. Krause, “The Genesis of Rocketborne Space Research” in Report of NRL Progress: Fifty Years of Science for the Navy and the Nation (ed. William M. Leak), pp. 46-48, Naval Research Laboratory, July 1973
Milton W. Rosen, The Viking Rocket Story, Harper & Brothers, 1955
M.W. Rosen, “The Pre-Vanguard Era at NRL”, in Report of NRL Progress: Fifty Years of Science for the Navy and the Nation (ed. William M. Leak), pp. 49-51, Naval Research Laboratory, July 1973
George P. Sutton, History of Liquid Propellant Rocket Engines, AIAA, 2003