By John McHale
NASA experts are upgrading the Space Shuttle orbiter's cockpit to a glass cockpit with commercial-off-the-shelf hardware such as liquid-crystal displays and PowerPC processors. Meanwhile, the International Space Station, still nearing completion, continues to rely on older custom-designed technology.
After nearly 20 years of flying with 1970s cockpit electronics, NASA's Space Shuttle orbiter is upgrading to a digital cockpit that reduces Shuttle crew workload and cuts down on human error. Once all four orbiters have completed the cockpit facelift in 2004, they will get new state-of-the-art computers to complement their new displays.
The upgrades signal that the world's most advanced spacecraft is finally starting to catch up with the technology on the average American's desktop computer. The latest high-speed Intel and AMD microprocessors are as many as five or six generations ahead of the computers that control the Space Shuttle.
Why the big generation gap? Well, the functions of a desktop computer currently require more processing power than flying a spacecraft into Earth's orbit and back. Even if NASA engineers wanted to keep up with the latest and greatest in computer technology, the cost of qualifying and testing each new piece of hardware for space would be astronomical — and only a fraction of the cost of requalifying the software.
More to the point, however, the old technology works just fine; in fact, it rarely if ever fails. Still, obsolescence is an issue, which makes commercial-off-the-shelf (COTS) hardware and the open-systems architectures attractive options for future upgrades.
COTS and open architectures
Space Shuttle designers have not been able to embrace COTS technology as fast as those in the military because of risk, says Elrick McHenry, manager of Space Shuttle program development at NASA's Johnson Space Center (JSC) in Houston. "We're still working with controlled explosions, so there has to be a heckuva a lot more testing and design review."
COTS products are also high-volume applications, which the Shuttle program is not, McHenry continues. There are only four orbiters in the Shuttle fleet — Atlantis, Columbia, Discovery, and Endeavour. Upgrading each is costly and takes years of design, test, and qualification before engineers can plug the new equipment in, McHenry explains.
To combat avionics obsolescence and reduce lifecycle costs, the upgrade, which begins in 2006, is going with an open architecture, says Mike Brieden, cockpit avionics upgrade project manager at JSC. This will enable NASA experts to chart upgrades for processors, boards, storage devices, and other hardware, he adds.
"We can't afford the time to custom-design hardware," Brieden says, adding that most of the custom work involves application software where "you get the most bang for your buck." The high risk of space operations mandates the reliability of custom-designed code; COTS application software is simply unacceptable, Brieden says.
Original code generation will continue to be custom work, but the Shuttle's smart cockpit upgrade will use the VxWorks real-time operating system (RTOS) from Wind River Systems in Alameda, Calif., Brieden says. VxWorks, which has a long track record in harsh military environments, is robust and reliable for space, Brieden says.
Until designers figure out a way to make software independent of hardware, software will continue to be the costliest part of any program because of its certification and testing — even in system upgrades, Brieden says.
Among the crown jewels of the Shuttle upgrade is its so-called "glass cockpit," which swaps out old mechanical and analog instruments and replaces them with software-driven multifunction liquid-crystal displays. Honeywell Space Systems in Phoenix is under a subcontract from Boeing in Seattle for the glass cockpit display upgrade called the Multifunction Electronic Display Subsystem (MEDS).
The NASA-JSC Space Shuttle Program Office in Seattle awarded Boeing the overall avionics upgrade contract. Boeing is responsible for the overall integration involved in the MEDS upgrade.
NASA officials through United Space Alliance (USA) in Houston contracted with Lockheed Martin Systems Integration's Aerospace division in Owego, N.Y., to upgrade the Integrated Data Processors in the MEDS cockpit to Command Data Processors under the smart cockpit part of the MEDS upgrade.
The first orbiter craft to receive the MEDS upgrade was the Space Shuttle Atlantis, in which the cockpit now has nine Honeywell MEDS or flat-panel displays in the forward flight deck and two displays in the aft flight deck.
Astronauts will use the displays to navigate and land the spacecraft, Honeywell officials say. The two MEDS displays in the aft flight deck support payload operations, NASA officials say. Overall there will be 140 displays including ground units, spares and those on board the four spacecraft, McHenry says.
"MEDS is a great step forward," says astronaut Jeffrey N. Williams, flight engineer on the STS-101 mission, which was the first flight of Atlantis with the MEDS cockpit. The MEDS digital environment will help the crew manage risk and cut down on human error, he adds.
Crew error causes far more than half of all aircraft accidents, which is why Williams calls the MEDS upgrade a "revolutionary step." With MEDS the crewmembers never lose situational awareness as they move throughout the cockpit, Williams says. They can call up flight data on any of the system's 11 displays, he adds.
Engineers at the Boeing Co. in Palmdale, Calif., performed the MEDS upgrades during each orbiter's major down period, which lasts about 10 to 12 months, McHenry says. During that time each spacecraft goes through major maintenance and modification, he adds.
Engineers gutted the old cockpit, which used three cathode ray tube (CRT) displays, on the Shuttles Atlantis and Columbia to take advantage of a graphics display, McHenry says. Engineers completed the Columbia upgrade earlier this year.
The MEDS upgrade replaces four CRT screens in the Shuttle cockpit, as well as mechanical gauges and instruments with the same kinds of full-color, anti-glare flat panel displays that Honeywell engineers installed on the Boeing 777 jetliner, says Steve Hopkins, program manger for MEDS at Honeywell. The glass cockpit is also similar to the cockpits on the U.S. Air Force F-15 jet fighter and U.S. Navy F/A-18 fighter-bomber, he adds.
MEDS uses ruggedized Philips liquid-crystal displays, which have error detection and correction (EDAC) circuitry to deal with single event upsets (SEUs) and latchups in the radiation environment of space, Hopkins says.
The Honeywell displays should last through 2020, yet engineers seek to mitigate obsolescence issues through planning and identifying suppliers who can provide long-term support, Hopkins says.
The current upgrade to a MEDS cockpit uses three Intel 386-based Integrated Data Processors (IDP) to generate information for the display, McHenry says. The IDPs receive their commands from the General Purpose Computer (GPC) for the cockpit avionics, he adds. The GPC is the AP-101S computer from Lockheed Martin Systems Integration-Owego. The IDP uses the old 386 microprocessor because certifying flight hardware takes a long time, McHenry says.
The old Shuttle cockpit used a hard-wired system with a dumb terminal to perform the roles that the IDP plays today, Brieden says. Before MEDS crewmembers had to go to different parts of the cockpit to get different information; now they can take the display with them wherever they go, Hopkins says. Astronauts will also be able to send and receive e-mail from home, he adds.
MEDS also has improved the Shuttle's caution and warning systems, McHenry says, which takes a lot of stress off the crew by providing them with the ability to move the data from one display to another, he adds. Unlike the mechanical gauges and dials of the old shuttle cockpit, MEDS displays provide NASA the flexibility to reconfigure cockpit graphics to fit the mission. This provides astronauts with more useable information with less effort.
The old electro-mechanical instruments of the old cockpit now appear as graphic images on any of the 11 MEDS screens, McHenry says. Shuttle crews also have easy access to vital information through the two- and three-dimensional color graphic and video capabilities, Honeywell officials say NASA experts have also developed color-coordinated standards for the new displays, McHenry says. Each type of data will have a different color representation on the display screen, he explains. For example, attitude data may be blue or another sensor's data may be designated by red, McHenry adds.
Besides reducing maintenance costs, MEDS saves a significant amount of weight and power and improves reliability and performance, NASA officials claim. The new MEDS displays will also offer common display hardware for multiple functions and expansion capability for future applications.
Honeywell engineers, under contract to Boeing, began designing the MEDS subsystem in 1992, and completed deliveries for all four orbiters in 1999. In 1995, Honeywell was contracted to design identical units for the International Space Station, which will enable astronauts to interface with a robotic arm on the station's Japanese module, Honeywell officials say.
Astronauts will be able to position the arm by viewing graphics and live video on the display unit, Honeywell officials say. In the past, astronauts could only control the arm by looking out the window.
The future smart cockpit upgrade will retain the MEDS displays but replace the three IDP processors with three CDP devices to improve graphics and processing, says Wayne Ordway, deputy project manager for the cockpit avionics upgrade at JSC. The smart cockpit is actually part of the whole MEDS upgrade, and the next step after furnishing the display technology, he adds.
Engineers at Lockheed Martin Systems Integration-Owego are developing four prototype hardware units to upgrade the Space Shuttle orbiter cockpit processor subsystems. They also will supply the command and display processor and the avionics ground equipment interface serializer for the new orbiter cockpit. The new hardware will interface with the company's AP-101S GPC and will provide functionality across the Shuttle's avionics suite, Lockheed Martin officials say.
The CDPs in the smart cockpit will correlate data from the spacecraft's sensors so the crew can concentrate on analyzing the information and making critical decisions instead of spending time gathering the data themselves, McHenry explains.
These prototype units will use the Lockheed Martin SP-103B PowerPC 750 computers with pre-existing Lockheed Martin-developed and application-specific modules, says Geoff Jones, the company's senior program manager for space programs. The new devices will have cache memory of 256k and will use static random access memory, JSC's Brieden says.
The components Lockheed Martin engineers are providing NASA for this upgrade do embrace commercial technology like the PowerPC, but there are some pedigree parts that cannot be COTS because of the stringent requirements necessary for space, Jones explains.
The SP-103 devices are designed to deal with diminished manufacturing sources and end-of-life issues, Jones says. Sometimes lifetime buys are necessary, but when upgrading hardware it is important to not change the software in order to keep costs down, he explains.
NASA radiation requirements for the cockpit upgrades prohibit latchups, but allow for EDAC circuitry to correct SEUs in the processor, Brieden says. A SEU occurring every 100 days is acceptable, because the EDAC can catch it, he adds.
Shuttle electronics designers must be alert to potential problems that may occur in radiation testing when a vendor changes his product line, Brieden points out. If the vendor can show no change that might affect radiation tolerance, NASA can still use the part, he explains. If not then they might have to look elsewhere and requalify and test all over again, he says.
The cockpit avionics upgrade contract win returned the Lockheed Martin Systems Integration-Owego operation to the Space Shuttle program for the first time since 1988, when company engineers upgraded the cockpit's GPC, the AP-101S. Owego engineers began supporting the Space Shuttle back in 1973, when IBM owned the company, Jones says. Its original contract was to build the onboard computer system for the spacecraft. In 1988 experts at Owego upgraded the original GPC, the AP-101B, to the AP-101S, which is the GPC on the shuttle today, Jones says.
"The AP-101S is rock-solid hardware projected to last to 2020," JSC's McHenry says. "It has a great MTBF [mean-time-between-failure] rate and we have no reason to modernize the processor." In 1998 the computer achieved an MTBF of 7.8 years and in 1999 improved that MTBF to 10.6 years, Lockheed Martin officials say. Current MTBF estimates were not available at press time.
There are four GPCs on the orbiter for redundancy, McHenry says. They can vote each other out in case of failures, he explains. The fifth GPC, a spare, was removed from the orbiters due to the outstanding performance of the main four devices — the spare was not needed, Lockheed Martin officials say.
Today most of the Shuttle's application software is written in Ada, however the software language on the original GPC, the AP-101B was called HAL, or high order application language, McHenry says. It was a tribute to the computer in the book and movie "2001, A Space Odyssey", and to IBM, Jones says. Arthur C. Clarke, the author of the book came up with the name by going back one place in the alphabet for each letter in IBM, which becomes HAL.
The Shuttle upgrade also includes Global Positioning System (GPS) receivers to provide accurate vehicle attitude and location data. This will reduce Shuttle costs by allowing the program to remove several ground stations that are becoming obsolete and costly to maintain.
GPS receivers along with some future navigation enhancements will enable precision Shuttle landings and provide technology applicable to many of NASA's future space programs, NASA officials say.
Other parts of the upgrade include the Integrated Communication System, which replaces the Shuttle's deployed Ku-band and S-band antennas with four common phased arrays that provide enhanced communication service for the Shuttle. The upgrade will reduce the number of potential failure modes in the Shuttle communication system, simplify on-orbit and flight operations, lower operating costs, and enable faster turnaround times at the launch site, NASA officials say.
International Space Station uses custom-designed electronics for guidance, navigation, and control
The International Space Station, while finally gaining momentum and attention with its recent change of crew, is still years away from open-architecture upgrades to its guidance, navigation, and control systems like the current program for the Space Shuttle's cockpit.
The station, much like the Shuttle when it first launched nearly 20 years ago, is using new technologies and designs on a scale never seen before. The lives of the astronauts working on the spacecraft and the billions of dollars invested require that it not fail. Therefore the majority of its electronics, while obsolescent compared to performance of a desktop PC, are thoroughly tested and qualified for human spaceflight.
The mission-critical custom electronics at the heart of the station's guidance, navigation, and control (GNC) system are Multiplexers/Demultiplexers (MDMs) computer units from Honeywell Space Systems in Phoenix. The two Station modules on orbit today, Unity node and Zarya, each have two Honeywell MDMs to control core systems.
In total, 52 Honeywell MDMs will serve as the Station's command and data handling system. MDMs control not only the Station's GNC system, but also its power, environmental controls, sensors, and effectors. MDMs also will manage payload operations and handle crucial ground-control instructions.
The MDM communicates with all the spacecraft's sensors and effectively gives commands to other components on the station based on sensor data, says Roger Hayes, manager of business development for the International Space Station at Honeywell Space Systems.
When NASA controllers at Johnson Space Flight Center (JSC) in Houston want to change the attitude of the Station, they communicate with the MDMs, which then order the Control Motion Gyroscopes (CMGs) to spin at rates as fast as 6,000 rpm to adjust the spacecraft, Hayes explains. The station's CMGs control the spacecraft's roll, pitch, and yaw.
The station can move via the propulsion thrusters on the Russian module, or with the CMGs on the U.S. module. However, most adjustments maneuvers are done with the CMGs, says Courtenay McMillan, one of the Station's attitude determination and control officers.
The main advantages of the CMGs over the propulsion system are a cleaner microgravity environment, less torque placed on the spacecraft "and it's cheaper to use the CMGs because in using thrusters, you have to count the cost of launching a Progress full of propellant. I don't know dollar signs for that, but it isn't cheap," McMillan says.
The attitude determination and control officer works at the GNC station at Station Flight Control at JSC and works in partnership with Russian controllers to manage the station's orientation with the onboard Motion Control Systems. This officer also plans and calculates future orientations and maneuvers for the Station. McMillan's station at mission control consists of a Compaq Alpha-based computer for mission-critical activities and a PC for other less critical functions.
Each MDM consists of an Intel 386 SX processor and a dual-channel MIL-STD 1553B databus card, Hayes says. The computer is a custom backplane based on the SX architecture, he adds. The MDMs also contain 64K of SRAM and 8 megabytes of DRAM, he adds.
The MDMs have discrete I/O card capability and interface directly with all of the box-level hardware. Each circuit card is removable. In orbit, the crew will be able to pull a box out of the rack and either replace individual cards or, if necessary, install a complete spare MDM.
A 1553 data bus moves data to important functions on the space station, while Ethernet moves data for payloads, McMillan says.
Currently, there are no obsolescence issues with the MDMs, Hayes says. However, Honeywell engineers are working on a Pentium replacement for the 386 device, but this upgrade is still in the early stages of research and development.
Controllers at JSC plan attitude together with their Russian colleagues in Moscow as the station orbits, she says. The software on the Station control systems is designed to work with gyroscopes and the Russian thrusters, she explains. The astronauts on board the spacecraft never actually control it, unless in a very rare instance where Station Flight Control cannot, McMillan says.
When maneuvering the Station via the CMGs, the U.S. GNC software controls four CMG assemblies that use large, spinning gyroscopic wheels to impart a constant torque on the Station, which holds the spacecraft in the desired attitude, NASA officials say. The inner and outer gimbals of the CMGs move imparting the torque to the Station. The U.S. GNC control software consists of the attitude control subsystem software and the CMG software located in the MDMs, NASA officials say.
Information critical to running an integrated U.S. GNC and Russian segment motion control system occurs between the command and control MDMs and the Russian segment central computer, NASA officials explain. Communication at this level is for handling contingency, moding, commanding, and monitoring of the GNC systems. This communication occurs on the 1553B databuses, NASA officials say.
When moving the space station with the CMGs, personnel at the GNC station at Station Flight Control "don't need to do much," McMillan says. The Station itself does not "move around a whole lot, it's not an airplane and not designed to fly like one," she continues. "Mostly we monitor and make sure all the CMGs [and other devices are doing their job]. When we do need to change things, we can command new attitudes, hand over to the Russian segment when necessary — and back, of course."
The GNC personnel also go through hours and hours of simulations to practice their responses in many different scenarios. In McMillan's case one of these scenarios involves CMG failure. "If one fails, no big deal," she says. "If two fails, slightly bigger deal, but we're still good. We need at least two CMGs to control attitude, so it'll take three CMG failures to trigger an automatic hand over to the Russian segment control system. Other failures can too, of course, like if we lose our GNC MDM. "
Collisions are possible because "there is a lot of stuff floating around up there," McMillan says. NASA experts are working with Russia to coordinate a collision avoidance system, she adds. Someone is always watching for debris that may collide with the space station, and if the situation is serious and the spacecraft needs to move quickly, control transfers to the thrusters, McMillan explains.
Control Motion Gyroscopes
Engineers at L-3 Communications Space and Navigation division, formerly Allied Signal, in Teterboro, N.J, made the CMGs and have also made gyroscopes for many other programs, McMillan says. The CMGs were part of the Z1 Truss package added to the station during the Space Shuttle's STS-92 mission last fall.
"A control moment gyroscope spins at a constant rate; two or more CMGs control a total angular momentum vector by "gimballing" the wheels — that is, rotating the spin axis," McMillan explains. "The more "parallel" the two (or more) spin axes, the bigger the total momentum vector.
"Think of it like coasting on your bike and then turning the handlebar," she continues. "The speed of the wheel isn't changing, just the angle of the axle — also known as the spin axis." The Hubble Space Telescope, she says, uses reaction wheels that control the total momentum vector by spinning wheels up or down — "like pedaling without turning the handlebars."
"By definition all gyros are spinning gyros," McMillan says. Laser gyros are for angular rate measurement only and cannot be used for control, she adds. "The Russian rate gyros are also mechanical (spinning wheels, in other words); the rate gyros planned for use on the U.S. segment are laser gyros."
Honeywell Space Systems engineers in Clearwater, Fla., also make the rate gyro assemblies that provide attitude reference, Hayes says. This data is especially critical for when the Space Shuttle docks with the Station.
McMillan says that sensors on the Russian segment provide the majority of the data needed for controlling the CMGs, and fall into three categories, which are:
- attitude determination, which includes star mappers, magnetometers, and infrared horizon sensors;
- angular rate determination, which includes gyroscopic sensors; and
- state vector determination, which includes Global Positioning System/Global Navigational Satellite System receivers — currently not in use.
The magnetometer determines attitude based on the spacecraft's position in relation to the Earth's magnetic field, McMillan says.
The MDMs are radiation tolerant not radiation-hardened, Hayes says. When dealing with human spaceflight, engineers must concern themselves with resistance to single-event upsets (SEUs) and latchups, but not total-dose resistance, he explains.
SEU occurrence depends on solar activity, Hayes says. The MDMs only see about two or three failures a year, he explains, and the error detection and correction circuitry fixes those errors. The MDMs also have redundancy outside the box, Hayes continues. There may be even dual or triple strings — two or three MDMs for one system — depending on the importance of the system, he explains. Most of all critical systems are dual or triple redundant, Hayes says.
COTS and the International Space Station
Some commercial-off-the-shelf (COTS) technology is aboard the station, Hayes says. A mechanical disk drive from the Raymond Engineering operations of Kaman Aerospace Corp. in Middletown Conn., is COTS, as are the displays on the Japanese Experiment Module, which are based on commercial devices in the Boeing 777 cockpit like the ones for the Space Shuttle cockpit upgrade, he explains.
The Raymond hard drive has a storage capacity of 300 megabytes with three circuit cards and Honeywell engineers are working on an upgrade to a solid-state device with one gigabyte of memory and comprising only three circuit cards and no mechanical drive, Hayes says.
The lack of COTS technology is mostly due to NASA's stringent requirements, he says. You also have to be careful when upgrading hardware, Hayes continues. While you may be getting a faster processor you will have to rewrite some software code and recertify it, which can be expensive, he explains.
Boeing engineers write the software on the station in the Ada programming language, Hayes says. Ada is still a reliable language and very suitable for space, he adds. The trick will eventually be to make the software independent of the hardware, Hayes says.
As for the possibility of upgrading the Station's 1553 technology to Fibre Channel, Hayes says it would be nice to upgrade to state-of-the-art technology but it is more prudent to wait until the whole station is up and working. Plus the wiring is already in place for 1553 and the technology works just fine for the Station's needs, he adds.
Led by the United States, the Station uses the scientific and technological resources of 16 nations — Canada, Japan, Russia, 11 nations of the European Space Agency, and Brazil.
More than four times as large as the Russian Mir space station, the completed International Space Station will have a mass of about 1 million pounds. It will measure about 360 feet across and 290 feet long, with almost an acre of solar panels to provide electrical power to six laboratories, NASA officials say.
The station is in an orbit with an altitude of 250 statute miles with an inclination of 51.6 degrees. This orbit enables the station to be reached by the launch vehicles of all the international partners to provide a robust capability for the delivery of crews and supplies. The orbit also provides excellent Earth observations with coverage of 85 percent of the globe and 95 percent of the population, NASA officials say. Already, about 500,000 pounds of station components have been built at factories around the world.
In all, over 100 different components will be joined together in orbit to construct the International Space Station with final assembly expected in 2004, NASA officials say.—J.M.
Motorola to develop software for space shuttle upgrade
Officials at United Space Alliance LLC, in Houston are asking experts at Motorola in Scottsdale, Ariz., to support cockpit avionics software design requirements for the upgrade of the Space Shuttle orbiter for NASA Johnson Space Center in Houston.
The $5 million contract, with options, eventually could be worth $30 million. United Space Alliance (USA), LLC, serves as NASA's prime contractor for the Space Shuttle program.
Developed by USA, the cockpit avionics upgrade is designed to improve the flight crew's insight and understanding of the Space Shuttle orbiter's health and status during operations and improve the content and quality of control mechanisms available to the crew. The upgrade requires the ability to efficiently manage software development and the ability to execute to Software Engineering Institute Capability Maturity Model Level Five, Motorola officials say.
"The development of the Space Shuttlesoftware upgrade is an unforgiving task which plays a critical part in the success of NASA's future space transportation objectives," says Dave Finkel, director of upgrades subcontract management for United Space Alliance, LLC. "Upgrading the Shuttle's cockpit avionics systems will reduce the workload of the crew and improves our astronauts' ability to respond to anomalies during critical times of a mission."
"Our software engineering organizations have demonstrated efficiency in software development to the Software Engineering Institute, a research center of the [U.S.] Department of Defense, at a very high maturity level," says Mark Fried, corporate vice president and general manager of Motorola's Integrated Information Systems Group. "We are fully committed to using our competency in streamlining the work process in the development of this sophisticated software for USA and for NASA."
The cockpit avionics upgrade project goal is to increase the crew's situational awareness while reducing their workload, NASA officials say. The Motorola software is considered a critical function to the vehicle, requiring the highest levels of software quality in order to ensure safe operations, NASA officials say. — J.M.
NASA research cuts effectively kill VentureStar space plane program
NASA leaders quit funding their X-33 spacecraft project in March, a move that effectively kills the initiative to build the so-called VentureStar single-stage spacecraft.
In a status report on their Space Launch Initiative (SLI), NASA announced the end of funding for the X-33 and X-34 space launch programs effective March 31, which ends the VentureStar program unless leaders of prime contractor Lockheed Martin Corp. of Bethesda, Md., decide to continue the project on their own.
"NASA determined that the benefits to be derived from flight testing these X-vehicles did not warrant the magnitude of government investment required and that SLI funds should be applied to higher priority needs," according to a NASA statement.
Government and industry reportedly own the X-33's hardware and software together. It is not clear yet how X-33 and VentureStar technology will go forward, if at all. The VentureStar was to have a unique avionics system, for example, in which mission and flight data would be processed in the same computer enclosure.
"We have gained a tremendous amount of knowledge from these X-programs, but one of the things we have learned is that our technology has not yet advanced to the point that we can successfully develop a new reusable launch vehicle that substantially improves safety, reliability, and affordability," says Art Stephenson, director of NASA's Marshall Space Flight Center in Huntsville, Ala. Marshall manages the SLI, X-33, and X-34 programs for NASA.
NASA began the X-33 program in 1996 as part of its Reusable Launch Vehicle program. It called for the demonstration of a subscale single-stage-to-orbit vehicle that would go from launch stand to orbit without using several different stages as the Saturn moon rocket did or dropping rocket motors and fuel tank like the Space Shuttle.
Using composite materials to reduce vehicle weight is one of the keys to successfully developing a single-stage-to-orbit launch vehicle, NASA officials say. In November 1999 the X-33's composite liquid hydrogen fuel tank failed during testing. An investigation into the cause of the failure revealed that composite technology was not mature enough for such a use.
Lockheed Martin proposed to complete development of the X-33 by replacing its two composite liquid hydrogen tanks with aluminum tanks. NASA agreed to permit them to compete for SLI funding to do so. But the benefits of testing the X-33 in flight did not justify the cost, NASA officials say.
NASA investment in the X-33 program totaled $912 million, staying within its 1996 budget projection for the program. Lockheed Martin originally committed to invest $212 million in the X-33, and during the life of the program increased that amount to $357 million.
The X-34 program also was initiated in 1996, to provide a low-cost technology test bed that would demonstrate a streamlined management approach with a rapid development schedule and limited testing. The prime contractor on the X-34 program was Orbital Sciences Corp. of Dulles, Va.
A joint NASA/Orbital Sciences review last year revealed the need to redefine the project's approach, scope, budget, and schedule, NASA officials say. To ensure safety and mission success of the X-34 it became necessary to increase government technical insight, hardware testing and integrated systems assessments. As a result, the projected cost of completing the X-34 program at an acceptable level of risk rose significantly above the planned budget.
NASA decided that such additional funding for X-34 risk reduction would have to be competed within the SLI evaluation process. As with X-33, NASA determined that the benefits to be derived from continuing the X-34 program did not justify the cost. — J.K.