How the Shuttle Orbiter Lost Its Jet Engines

The McDonnell Douglas design with a large flyback booster

As design work by various aerospace companies began on the Space Shuttle program in the late 1960s, it was a given that the Orbiter would have its own jet engines. Having its own air breathing engines offered three advantages- they would allow atmospheric flight testing much like any other aircraft was tested and pilots could practice landings in the run up to an orbital mission. The engines also facilitated ferry flights, repositioning the Orbiter amongst various facilities (landing, launch, overhaul, etc.). Having its own jet engine propulsion also gave the Orbiter cross range capability upon return from orbit. Some designers envisioned the Orbiter rendezvousing with a tanker for additional jet fuel. But in the ascent and in orbit, jet engines and fuel for those engines was dead weight that subtracted from potential payload. Even if designers went with an Orbiter design that was unpowered on its landing, the 1970 and 1971 design studies prominently featured a fully reusable two stage Space Shuttle with a big flyback booster that would have to have its own jet engines. Some of the designs for the flyback booster were massive with a need for as many as twelve jet engines. Soon the design of the flyback booster itself began to take on technical challenges that rivaled that of the Orbiter design itself. The weight of up to twelve jet engines and the necessary jet fuel cut into the payload of liquid hydrogen and liquid oxygen for the booster's rocket engines. Many of the flyback booster designs would need approximately 150,000 lbs of jet fuel (for comparison, a Boeing 777-200ER has a fuel capacity of roughly 300,000 lbs). Consideration was then given to using liquid hydrogen as fuel for the jet engines which would cut out the need for jet fuel tanks. In June 1970, NASA issued contracts to GE to study the feasibility of using liquid hydrogen in the F101 engine being developed for the B-1 bomber. Pratt and Whitney also got a similar contract to study the use of liquid hydrogen fuel in the F401 engine, the planned naval derivative of the USAF's F100 engine planned for the F-15 Eagle. Both companies showed that liquid hydrogen fueled jet engines saved about 2500 lbs of weight per jet engine compared to conventionally-fueled jet engines. The weight savings was modest at best. 

A typical high-key to low-key unpowered approach to landing

At the same time these studies were going on on how to save weight with Orbiter and flyback booster-mounted jet engines, with NASA there was a group at the Flight Research Center at Edwards AFB where unpowered landings were routine for many high speed research aircraft going back to the X-1 (the X-15 program being the most recent one at the time) and the graduates of the co-located Aerospace Research Pilot School had as a requirement that students demonstrate proficiency in unpowered landings using the school's Lockheed F-104 Starfighters which were throttled down to idle for the practice sessions. Even more demanding were the unpowered landings made by the lifting body program aircraft that lacked wings and derived their lift from their tubby fuselage designs. Regardless of what sort of aircraft was used, USAF test pilots and the NASA-FRC pilots used what was called "energy management" where they traded altitude for airspeed on the descent and used turns to bleed off speed in preparation for final approach. The first step in unpowered landings was the arrival at the "high key" which was high above the touchdown point. From the high key, a gradual 180 degree turn was made that allowed speed reduction and descent to the "low key" which was usually abeam the touchdown point. From the low key, the turn continued allowing more speed to bleed off and the descent to continue until lined up for final approach. If at any point the speed was excessive, speed brakes or gentle S-turns could be used to get down to the necessary airspeed. The lifting body pilots found that on final approach, diving at the runway touchdown point 15 degrees or more improved their accuracy as the speed improved the stability and the speedbrakes could be used to moderate the speed build up on final approach. An assessment by one of the experienced lifting body pilots in September 1970 showed that in 30 landings on a 10,000 foot runway from altitudes as high as 90,000 feet and speeds as high as Mach 2, the dispersion of the landing points was only 250 feet. 

However, the astronaut office in Houston at the Manned Spaceflight Center headed by Deke Slayton felt that unpowered landings for the Orbiter were too risky. Slayton was concerned that the test pilots were more proficient at unpowered landings than his astronauts would be, especially if they were returning from a 7-10 day orbital mission. The astronauts' views carried considerable weight for good reason and it took the USAF to swing the design work in favor of unpowered landings. 

I had posted previously that the Space Shuttle program's development phase was taking place during a period of budget austerity. One of the keys to navigating the budgetary climate of the day was to be sure to secure as much political support as possible since Congress determined the program budget. But in 1970 the program had some close calls, narrowly avoiding funding cuts in both the House and Senate. The Air Force offered to lend its support as it saw opportunity in the Shuttle program to launch heavy reconnaissance satellites. But NASA had baselined the Orbiter design at the time with a 25,000 lb payload to orbit. The USAF wanted to put its heavy reconnaissance satellites into polar orbit and the Orbiter needed a payload capacity of 40,000 lbs. That much payload weight into polar orbit (and unable to take advantage of the Earth's rotation for additional boost) was equal to a 65,000 lb payload launched for the Kennedy Space Center. NASA informed the USAF that the payload had to be baselined at 25,000 lbs due to the weight of the jet engines and their fuel. But it was apparent from the Congressional battles that NASA needed a strong ally like the USAF, so the jet engines were dropped from the Orbiter design and that allowed the payload capacity to orbit to meet the USAF requirements. 

The idea of onboard jet engines didn't end, though. NASA shifted towards the idea of removable kit that could be used for flight testing, ferry flights, and for return from orbit if the payload wasn't maxed out. This also coincided with the 1971-1972 time frame when the flyback booster was dropped as too much of a technical risk and the Space Shuttle began to look more like its final design- an Orbiter with an external tank and solid rocket boosters in what was called the TAOS configuration- Thrust Assisted Orbiter Shuttle. The significant weight savings by going to a TAOS configuration also helped cut development risk as there was a considerable amount of experience already with solid rocket boosters and large external tank structures to hold cryogenic fuels. 

The test pilots at NASA-FRC persisted in their opinion that jet engines were completely unnecessary in the Orbiter design. They had their long experience of over 10,000 unpowered landings since the X-1 program as their proof, but the astronauts insisted that the Orbiter was a much bigger aircraft than many of the X-planes. Another round of tests then were held by NASA-FRC, this time using their B-52 Stratofortress carrier aircraft. Set up in a high drag configuration with the engines at idle, pilots successfully and accurately landed the B-52. NASA-FRC then got some lifting body pilots who had never flown anything as big as the B-52 and had them fly the bomber through a simulated unpowered landing using energy management. They were able to land successfully and when the same pilots were asked to land the B-52 using a conventional powered low angle approach, none of them were able to do so. The test pilots the FRC even brought into two United Airlines pilots to fly the B-52 in simulated unpowered landings and they had no issue doing so, reporting that such landings were much easier than conventional landings. The test pilots then followed up the B-52 tests with the same tests using NASA's Convair 990 which could simulate the Orbiter aerodynamics on landing. 

The final iteration of a jet-engine powered Shuttle Orbiter (from the Dennis Jenkins book)

NASA finally got agreement to go to exclusively unpowered landings on return from orbit for the Shuttle Orbiter, but the jet engines still didn't go away. At the time of Rockwell's award in 1972, the Orbiter design featured two engines that deployed from the payload bay and two more engines that could be mounted on struts. Less than six months later, the Orbiter design dropped the internally mounted jet engines completely and they were to be mounted as a kit on the flat underside when needed for flight testing and ferry missions. It finally took the ferry range to kill the engines completely from the Orbiter design. The Orbiter was similar in size to a Douglas DC-9 but had twice the weight. It had a lot of drag since it wasn't optimized for atmospheric flight and the delta wing was highly loaded. With five jet engines mounted in pods on the underside and tank of jet fuel in the payload bay, the Orbiter had a ferry range of only 500 miles. With Space Shuttle sites across the nation and contingency fields overseas, a 500 mile range was simply unacceptable. NASA looked at aerial refueling during ferry, but this added complexity to a design that was already experiencing cost overruns. In February 1974, NASA deleted the jet engine requirement completely. As a result, both for flight testing and ferry flights, the Orbiter would need a carrier aircraft, but fortunately that was a lot more straightforward a development process!

The Buran analog with its four AL-31 jet engine nacelles

Interestingly in the Russian Buran Shuttle program, there was an aerodynamic test analog designed OK-GLI that made 25 atmospheric test flights with four Lyulka/Saturn AL-31 jet engines mounted in nacelles in the aft fuselage. A fuel tank sat in the payload bay. The AL-31 is the jet engine that is used on the Sukhoi Su-27 Flanker. Nine taxi tests and 25 test flights were made using the Buran analog from December 1984 to December 1989. The engines were used to takeoff and then were throttled back on the descent to landing. All of the flight testing took place at the Baikonur Cosmodrome. The operational Buran, however, would not have jet engines at all and the Antonov An-225 Myria was developed as the carrier aircraft to ferry the Buran orbiter. 

Source: Development of the Space Shuttle 1972-1981: History of the Space Shuttle, Volume Two by T.A. Heppenheimer. Smithsonian Institution Press, 2002, pp85-92. Space Shuttle: The History of the National Space Transportation System- The First 100 Missions by Dennis Jenkins. Specialty Press, 2008, pp187-192. Photos: NASA, Wikipedia, Dennis Jenkins. 

Flight Testing on the Cheap: The Ground-Breaking M2F1 Lifting Body

In the early 1960s NASA's research work into lifting body designs was in full swing with the X-23/PRIME program testing lifting bodies in the re-entry regime from orbit. I had previously blogged about the PRIME program and the data that it provided the aerospace industry. But while the unmanned X-23 program answered the question of a lifting body's flight characteristics at the high speeds encountered during re-entry, questions still remained about the suitability of a lifting body in subsonic flight- would a lifting body design be controllable and stable in the low speed flight regime upon return to Earth of a reusable space vehicle? At the time, Dale Reed, an engineer at NASA's Flight Research Center (FRC) at Edwards AFB, had been following the lifting body research in his free time. Being an avid remote-control aircraft model builder, Reed built a 24-inch flying model of what was designated the M2 shape- the M1 shape was the product of research at NASA Ames which started out as a 13-degree half cone with a rounded nose. The M1 shape was refined into the M2 shape which added twin vertical fins, a canopy for a pilot, and horizontal control surfaces. This M2 shape was what Dale Reed modeled in his free time and dropped it from a 60-inch remote control "mothership" while his wife Donna would film the M2 model flight with an 8mm camera. He later showed the films to the director of the FRC, Paul Bikle, and Alfred Eggers, one of the pioneering researchers on lifting body work. They were suitably impressed with Reed's work that in November 1962 Bikle authorized a six-month program paid out of the FRC's discretionary fund to build a manned glider version of Reed's M2 model. 

The M2 glider had a steel tubular frame to which was attached the pilot's seat and landing gear (taken from a Cessna 150)- the exterior would be a light wooden shell this way different shapes could be tested. As it was, work on the lifting bodies within NASA would eventually settle on the M2 shape as the most ideal and the glider was designated the M2-F1. Space was set aside in the main hangar at the FRC which was cordoned off and nicknamed "Wright's Bicycle Shop". A local glider builder was contracted to help construct the outer shell while NASA engineers and technicians, many of whom were members of the Experimental Aircraft Association (EAA), lent their time and talents to the construction of the M2-F1. As many of them did it in the free time, it ended up costing NASA only $30,000 to build the M2-F1 when it was finished in early 1963! One of the reasons Paul Bikle funded the program out of FRC's discretionary fund was that he feared NASA headquarters in Washington would disapprove, or worse, take control of the project and make it excessively complex and expensive. 

The finished M2-F1 was 20 feet long, 10 feet high, and 14 feet wide. At the aft ends of the blunt half cone it had twin fins that mounted horizontal control surfaces nicknamed "elephant ears". The main body itself had two flaps at the end that were for trimming the glider in flight. Complete with the pilot, the M2-F1 weighed only 1,138 pounds. Later in the flight test program a Weber zero-zero ejection seat was fitted, but for the first set of flights, no ejection seat was used. By March 1963 the full scale wind tunnel tests at NASA Ames were completed with encouraging results and back at Edwards AFB tow runs were made with various vehicles pulling the M2-F1, the first captive flight taking place 5 April 1963. For these initial flights, the M2-F1 was suspended below a Rogallo wing as the vehicles weren't fast enough to pull the M2-F1 to its rotation speed. 

The engineers wanted to a faster vehicle to get the M2-F1 airborne before pulling it aloft behind a tow aircraft to higher altitudes. A 1963 Pontiac Catalina convertible was procured and stripped down. It had a 455-cubic inch engine, four barrel carburetor, and was a five speed stick shift. The engineers enlisted the help of famed California hot-rodder Bill Straup in Long Beach to replace the street tires with racing tires and to tune and tweak the engine for maximum performance. Rollbars and radio equipment were installed and the right hand front passenger seat was turned around to face aft. Adorned with the suitable high visibility markings and NASA logos, the FRC's hot rod pulled the M2-F1 for the first time airborne in June 1963. Satisfied with the near-ground/low altitude peformance, a Douglas R4D (Navy version of the DC-3/C-47) pulled the M2-F1 glider aloft for the world's first piloted lifting body free flight on 16 August 1963 when NASA test pilot Milt Thompson cast off from the tow plane at 10,000 feet and landed smoothly onto Rogers Dry Lake below. Six other test pilots flew the M2-F1 glider after casting off from the R4D towplane- Bill Dana, Capt. Jerauld Gentry, Don Mallick, Bruce Peterson, Donald Sorlie, and Col. Chuck Yeager. Bruce Peterson and Milt Thompson flew the majority of the M2-F1 flights. The M2-F1 eventually got a proper ejection seat and a small rocket was added which was fired before landing to assist with the pre-landing flare. Eventually modifications and test equipment required the M2-F1 to trade in its Cessna 150 landing gear for the gear of a Cessna 180. 

The last M2-F1 flight took place on 16 August 1966 after just over 100 flights and 400 ground tows. It was donated to the Smithsonian Institution and was restored years later and returned to NASA Dryden (what the FRC is today known as) for display. Proving that a lifting body shape could be safely flown at subsonic speeds, NASA proceeded to develop the M2-F2, a heavier, alumimum, rocket powered version that was dropped from the NB-52A and was capable of supersonic flight. Plans were to build two M2-F2s, but eventually one M2-F2 was built to supersede the M2-F1 and a competing lifting body design was built by Northrop as the HL-10. As an interesting bit of television trivia, the crash depicted in the opening of the "Six Million Dollar Man" is Bruce Peterson crashing in the M2-F2 on 10 May 1967, which he did survive. 

Source: Space Shuttle: The History of the National Space Transportation System- The First 100 Missions by Dennis R. Jenkins. Specialty Press, 2008, p36-38.

Warship in the Desert: The Muroc Maru

At the instruction of General Henry "Hap" Arnold, in 1933 Rogers Dry Lake in the Antelope Valley north of Los Angeles was chosen as a gunnery and bombing training range for the expanding US Army Air Forces. Set up six miles west of the town of Muroc by personnel from March Army Air Field in Riverside, a series of practice targets were set up throughout the dry lake bed site which included simulated structures, the outlines of naval ships for high level bombing practice, and even a track for a moving target. Retired elderly Keystone biplane bombers were set up as well on a simulated airfield. Two years after being established, a second site was set up on the dry lake bed just one mile north of the town of Muroc.

Initially needing only a caretaker group of 13 Army personnel, the Muroc bombing range operated as a branch of March Army Air Field and the size of visiting squadrons would result in groups of personnel of varying sizes temporarily taking up residence at Muroc. Not only bomber and attack squadrons passed through the range but fighter squadrons as well for ground strafing exercises. By the summer of 1941 the base had expanded considerably in anticipation of the coming war. By 1942 over 40,000 personnel were stationed at Muroc at any given time and this resulted in Muroc being established as an air base in its own right as Muroc AAF independent of March AAF.

One of the most unique targets at Muroc AAF was completed in March 1943 and was designated as "AAF Temporary Building (Target) T-799, a full-sized wooden mockup of a Japanese Atago-class heavy cruiser. It was made of wood and covered with chicken wire and tar paper. Over the tar paper it was covered with shredded chicken feathers to give the appearance of a solid warship. Constructed at the south end of Rogers Dry Lake, the target got the nickname "Muroc Maru" and cost the USAAF approximately $35,000, a hefty sum in those days for something to be used for target practice.

Sand berms were constructed around the Muroc Maru to simulate the bow wave and wake. Bomber crews practiced low-level skip bombing against the Muroc Maru. On particularly hot days, pilots reported that the mirage effect would give the appearance of the Muroc Maru sailing at sea. Due to the strong desert winds and temperature extremes, the chicken feather covering of the target wasn't as durable as the wood and chicken wire structure.

Flight testing of new aircraft began in 1942 at Muroc. Training activities ended at Muroc AAF in 1946 and with an independent US Air Force, it became Muroc AFB in 1948. The following year the base was renamed Edwards AFB.

As for the Muroc Maru, it was a declared a flight hazard and dismantled in 1950. Army engineers responsible for the dismantling weren't thrilled about the extent of unexploded ordinance in the structure of the target. Some of the sand berms remain at the site to this day as well as millions of nails and metal staples from the 1950 disassembly.

Source: X-Plane Crashes: Exploring Experimental Rocket Plane, and Spycraft Incidents, Accidents, and Crash Sites by Peter W. Merlin and Tony Moore. Specialty Press, 2008, p13-14.

Jet engine technology in the late 1940s was still immature enough that a jet-engined fighter often still lacked comparable range to a piston-engined fighter. As a result, both the USAF and US Navy saw the turboprop as the potential answer- the USAF hoped a turboprop fighter would offer a longer range and the US Navy was interested in the power of turboprop fighter to make takeoff from a carrier deck possible without a catapult. To evaluate the possibility, both branches agreed to fund an experimental turboprop fighter based on Republic's AP-46 design. The USAF would buy three and the US Navy would acquire a single example.

By 1951 the arrangement had changed to with only the USAF getting two AP-46 aircraft designated XF-84H. Although a development of the F-84F Thunderstreak, the XF-84H only had wing and canopy of the Thunderstreak with an all new fuselage and tail unit. A 5,580-horsepower Allison XT40 turboprop that drove via an 18-foot shaft a 12-foot three-bladed constant speed propeller. The Aeroproducts prop rotated at 3,000 rpm which meant that the blade tips traveled at Mach 1.18 and as such, created an immense amount of noise, leading to the XF-84H's name, the Thunderscreech. As the prop was constant speed, thrust was effected by adjusting the pitch of the prop.

With its first flight at Edwards AFB on 22 July 1954, it was immediately apparent the XF-84H had some serious flaws. Maximum deflection of the rudder was insufficient to counter the massive torque and p-factor of the supersonic prop. In addition, there were constant hydraulic problems, the elevator was limited in its pitch authority and the prop gearbox kept overheating.

But most famously, the XF-84H Thunderscreech was noisy. The supersonic prop blades were so loud that residents over 20 miles away from Edwards AFB filed noise complaints with the USAF! When the aircraft first taxied out under its own power and took off, the control tower crew often had to hide under their consoles shielding their ears. Any ground crew standing nearby often felt nausea and dizziness even with ear protection due to the Mach shockwaves from the prop blades.

As a result, Republic was banned from doing run ups on the flightline and it would have to be towed 7 miles away from the main ramp behind a ridge called "Rocket Hill" as it had rocket engine test stands in the area.

The second Thunderscreech arrived in May 1956 and was temporarily bailed to the US Navy in support of the VTOL tail sitting convoy fighter program as both the Lockheed XFV-1 and Convair XFY-1 used the same Allison XT40 turboprop engine. The last Thunderscreech flight was made on 9 October 1956 with cancellation following, much to the relief of the pilots, ground crew and most everyone at Edwards AFB!

Source: International Air Power Review, Volume 24 (Summer 2002). AIRtime Publishing, 2002. "Warplane Classic: Republic F-84, Thunderjet, Thundersteak, and Thunderflash" by David Willis, p124.