Space Shuttle Design and Lessons Learned

UNCLASSIFIED

Space Shuttle Design and Lessons Learned Dr. Charles J. Camarda NASA Headquarters 300 E Street SW Washington, D.C. UNITED STATES of AMERICA [email protected]

ABSTRACT The Space Shuttle was an amazing and unique spacecraft that transported crewmembers, supplies, equipment, experiments, and large payloads both to and from low Earth orbit (LEO). It served multiple functions to satisfy diverse military, civilian, industry, and scientific requirements including: facilitating satellite launch, servicing, and return to Earth; serving as a science platform to conduct experiments in a micro-gravity space environment; providing large payload capability to LEO and moderate capability to geostationary Earth orbit (GEO); and serving as a low-cost space transportation system to help build the International Space Station (ISS). Space Shuttle was hoped to usher in a new era of safe, low-cost access to space, which would enable effective commercial and private utilization of space for everyone. This paper summarizes the development of the Space Shuttle, beginning with the exploration of hypersonic flight and reusable space vehicles and ending with the accomplishments of the engineers, scientists, astronauts, and program managers, whose achievements led to the program’s development and operation. The paper concludes with a review of some of the successes, failures, and lessons learned throughout the life of this incredible program.

1.0 INTRODUCTION The Space Shuttle was a unique, partially reusable space transportation system that was the culmination of over 40 years of rigorous and methodical research in analysing, designing, and testing vehicles capable of flying at hypersonic speeds. The development process followed a stepwise, building-block approach that matured with modelling and analysis, and which was firmly grounded by experimental testing and subsequent correlation with analysis. Members of the research community involved in the program embraced the importance of failure, acknowledging it as a necessary condition for the advancement of knowledge and an understanding of what was not yet known. This paper will discuss the early supersonic and hypersonic research and development programs that contributed to the design of the Space Shuttle and identify the underlying themes, processes and lessons learned, that were critical to its successful development and safe operation. In particular, special attention will be paid to the critical technical issues which had to be resolved such as: understanding hypersonic aerodynamics, high-speed stability and control; aerothermodynamics and associated heating; high-temperature materials and structures issues; and biomedical and human factors concerns. An understanding of the issues related to increasing speed regimes and altitudes was gained by a series of rocket and airplane tests of full-scale flight test articles, wind tunnel models and tests; and laboratory experiments which were systematically correlated and validated. As in any research program, test failures were necessary, to explore the unknowns and expand the flight envelope. Failures served as lessons learned and were the impetus for rapid discovery and learning which accelerated the knowledge and technology development for the Space Shuttle Program.

STO-AVT-234-VKI

PAPER NBR - 3

UNCLASSIFIED

ENTER CLASSIFICATION

Space Shuttle Design and Lessons Learned

2.0 UNDERLYING AND RECURRING THEMES DURING THE SPACE SHUTTLE DEVELOPMENT JOURNEY This paper will cover several recurring themes that are common to major programs — successful or unsuccessful. They will be highlighted throughout the analysis, design, fabrication, and test of US experimental flight test programs such as the X-series, as well as the laboratory, wind tunnel, and hypervelocity test programs. A subset of the many themes [1] and lessons learned from these programs will be discussed in the following sections.

2.1

Analysis, Design and Correlation with Test

A simple model of the general research and design process is shown in Figure 1 [2]. It begins with an observation of a physical event or an idea/concept, which we attempt to understand, using analytical or numerical means. We then evaluate our representation of that observation through a test or experiment. More often than not, our ability to predict the observed behavior of the experiment falls short; we “fail”. The error could be attributed to: 1. Our experimental representation of the “real” observation (e.g., initial conditions, boundary conditions, physical properties, etc.) or 2. Our simplified model (e.g., simplifying assumptions, numerical model, etc.). We iterate between experiment and analysis until we understand the discrepancies and are able to correlate our analytical representation of the behavior we observe in the laboratory to our predicted behavior to within a desired accuracy. It is important that during this process we are able to test concepts to failure. True understanding of the problem is affirmed when we can anticipate every conceivable failure mechanism and accurately predict when and how a failure will occur. With a true understanding of the physics of a problem we can then proceed to step into the world of design. Today, we use automated methods to rapidly and systematically vary design variables to explore the design space; identify a design that satisfies all constraints; and produce an “optimum” solution to our objective function (e.g., minimum mass, maximum payload-to-orbit, least cost, etc.). This process continues at each stage of a stepwise, building-block approach toward understanding the behavior and developing the concept(s). Throughout the process we fail and, consequently, repeatedly discover or learn that we have exceeded our understanding of the problem by moving beyond the bounds of our prior assumptions. It is in this region of the design space where we encounter an unanticipated “failure mechanism.” This process is iterative, and through it we develop a much better understanding of a research or design challenge. More importantly, we develop an understanding of the limitations in predicting behavior. An example of the stepwise, building block approach is the development of a reusable cryogenic tank concept for the single-stage to-orbit (SSTO) X-33 vehicle [3]. (A schematic of the approach used in the development of this concept for the SSTO X-33 vehicle is shown in Figure 2.) A series of thermal, structural, and operational test specimens and tests are strategically planned to address a particular failure mechanism and/or operational procedure early in the development process early in the design cycle, before performing a full-scale cryogenic tank test.

PAPER NBR - 3

STO-AVT-234-VKI




Figure 1. – Model of the research and design process [2]. The reusable cryogenic tank concept development follows this process in order to understand basic principles using simple models and experiments. Doing so validates that a simple analytical model is sufficient to accurately describe the behaviors witnessed in an experiment. Larger, more complex tests explore how well our physical understanding of a more complete and full-scale representation of the problem can be modelled and tested. These tests may include interfaces between multiple components, attachments, manufacturing details, etc. The closer the test article approaches the true embodiment and operational use of the concept, the more rigorous the analysis must be in order to predict behavior, and performance, and failure.

Figure 2. – Example of a stepwise-building-block approach in the design of a reusable cryogenic-tank for the single-stage-to-orbit (SSTO) X-33 vehicle. The initial or conceptual design phase of a project is the most critical part of its life cycle. It is the best time to positively affect the outcome (i.e., success) of a project. As shown in Figure 3, during the early part of the design cycle for most programs, the total life cycle cost (LCC) committed during the conceptual design phase is often insignificant (~1%). However, the total LCC incurred is about 70%. This tells us that down-selection to a design too early in the product development cycle can cause

STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned significant penalties downstream and may even result in an infeasible or intractable design. Contributing factors include insufficient or inappropriate analysis, test, and evaluation, thus committing a program to a sub-optimal concept. For example, the improper assessment of the debris threat to the Space Shuttle Orbiter’s Thermal Protection System (TPS) resulted in unanticipated post-flight system inspections, repairs and replacements (R&R), and significant operational expenses.

Figure 3. – Leverage in the Development Process (Image reprinted with permission of Prof. Olivier de Weck, MIT)

2.2

Systems Engineering and Critical Thinking – “Zooming Out and Zooming In”

Every problem is a multidisciplinary problem. For instance, the Apollo Program’s vehicle systems were so complex and interdependent that a new field of engineering was necessary to decompose the problem into digestible pieces and re-integrate them to evaluate overall behavior of the system. The practice of systems engineering (or systems thinking) emerged as a result, and is often used to attend to each discipline and its interfaces. The definition of “systems engineering” can vary depending on the phase of the design process. For example, at the conceptual level of the design process, a “systems engineer” is more focused on exploring many possible concepts by opening up the design space and relaxing design constraints. This focus shifts to hardware development in the later stages of the project life cycle when configuration management, process development, and quality control take higher priority. In essence, a systems engineer must be able to “zoom out” and understanding the big picture, and “zoom in” to increase the fidelity of the analysis, when needed, to fully understand the system’s behavior. Costly failure(s) can result if the smallest element or component is not considered within the context of the entire system.

2.3

The Importance of Failure

The importance of testing to and learning from failure cannot be understated as a fundamental part of research and development. Dr. Jack Matson observed that an innovative environment is one in which engineers and scientists are allowed to practice “intelligent, fast failure” through iterative experimentation [8]. It is through failure that we learn most effectively. Discouraging it hinders discovery, exploration, and innovation. Strings of successes can mask insidious failures that our simple models of behavior cannot predict. Success combined with a “can do” spirit can lead to arrogance. This can perpetuate an “overconfidence bias” or confirmation bias [5], resulting in the subjective interpretation of data to confirm what we want to be true rather than what is actually true. The Columbia Accident Investigation Board

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned (CAIB) Report [4] discusses this phenomenon in great detail in relation to organizational biases and culture that contributed to the loss of the Space Shuttle Columbia and her crew on February 1, 2003. Professor Henry Petroski of Duke University has written many books on the importance of failure and the limitations in understanding structural analysis of bridge and building design [6, 7]. His writing highlights how… simple assumptions in our analytical models used to describe observed behaviors can be wrong. However, not all failures are created equal. Author Amy Edmundson describes a “spectrum of failure” whereby failures can range from “praiseworthy” to “blameworthy” [8]. ***** The themes discussed in this paper above will reappear in the sections that follow. They serve as a reminder to be vigilant in our design processes, to learn the lessons of the past, to acknowledge and even praise failure, and to be humbled by our successes—we don’t always know how close we are to the precipice.

3.0 HYPERSONIC RESEARCH PRE-SPACE SHUTTLE During the 1920s and 1930s, German engineer Eugen Sänger envisioned vehicles moving at hypersonic speeds using rocket propulsion. One of his early design ideas was a two-stage “Rocket Bomber” called Silbervogel (Silverbird). This suborbital bomber aircraft was designed to be propelled by a rocket-powered sled up to 500 mph (800 km/h) and, once airborne, fire its rocket engine to climb to an altitude of 90 miles (145 km) [11]. Once it had exhausted its fuel, it would descend toward Earth, skip off the upper stratosphere several times to extend its range, and descend through the atmosphere deliver its bomb and land. While Sänger’s concept never went beyond the design phase, the development and testing of hypersonic vehicles ramped up when V-2 (A4) technology of the German scientists at Peenemünde was selected for use during World War II (WWII). This work was later adopted by the United States (US) and that would later serve as the foundation for future hypersonic vehicle development by the National Advisory Committee for Aeronautics (NACA) which later would become NASA. Critical to success was a partnership between NACA and the Department of Defense (DOD), which enabled the rapid development of laboratories and facilities (e.g., wind tunnels, arc jets, shock tunnels/tubes, free-flight tunnels, etc.) to validate analytical methods. This experimental data shaped our understanding of the behavior of vehicles at hypersonic velocities and helped optimize their design.

3.1

Early Rocket Work – Sänger to von Braun

Elements of the reusable two-stage Silverbird concept (see Fig. 4) served as one of several ideas that inspired the development of the Space Shuttle and later visions for hypersonic space planes such as the National Aero-Space Plane [11]. Early ballistic missile work progressed in the US under the leadership of Wernher von Braun and the team of German scientists and engineers from Peenemünde, who were acquired through Operation Paperclip, a clandestine program designed to recruit German scientists in the aftermath of WWII. American and German engineers and scientists collaborated to use V-2 rocket technology and its derivatives to achieve the first hypersonic speeds during a Bumper 5 rocket test from White Sands Missile Range (Figure 5) on February 24, 1949, attaining a maximum speed of 5,150 mph (8,288 km/hr.).

STO-AVT-234-VKI

PAPER NBR - 3




Figure 4. – Eugen Sänger’s designs for a rocket aircraft, the Silbervogel (“Silverbird”), and the Amerika-Bomber, which used a skipping entry [10]. NACA and German engineers continued to advance the state of rocket science, using the processes described in Figure 1. During this time, the rocket development program relied primarily on full-scale testing, as there was little confidence in wind tunnel capabilities at high speeds. Facilities had to be developed at a frenetic pace to correlate and validate data from full-scale flight tests [11]. There were many failures throughout these tests (http://www.youtube.com/watch?v=Ii7uwp1SRIM). Engineers encountered unknowns in aircraft stability and control; aerodynamic heating; aeroelastic instabilities (e.g., flutter); and thermal stress. These failures shaped the collective understanding and helped to develop technology that addressed these elements in areas such as flow physics, heat transfer, and materials science.

Figure 5. – V-2 (A-4), A-4B/A-9 and Bumper 5 Rocket Tests post WW II by the United States [10]. Early rocket configurations explored multiple stages and some, such as the A4-B had winged stages (Fig. 5) to significantly increase its range and effectiveness for military use. While the early conceptual studies and designs of Sänger were visionary, his lack of understanding about the hypersonic flight regime and associated aerodynamic heating would have prevented their success. His vision did, however, inspire later ideas for the Space Shuttle and even later visions for hypersonic space planes such as the National AeroSpace Plane (NASP) [11].

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned The military played a significant role in the advancement of aeronautics in the United States after WWI, and in high-speed, supersonic/hypersonic aerodynamics/aerothermodynamics, following WWII to support advanced weapons and weapon delivery systems. NACA was a vital partner with the DOD during this period; rapidly developing the laboratories and facilities (wind tunnels, arc jets, shock tunnels/tubes, freeflight tunnels, etc.) and validated analytical methods to help understand behaviors at these elevated velocities and optimize designs (the design ellipse of figure 1). This partnership was sustained during the early planning stages of the Shuttle Program and was critical in setting several of the Shuttle design requirements.

3.2

Hypersonic Airplanes and Lifting Bodies – “The Thermal Barrier”

The transition from supersonic flight to hypersonic flight brought with it a new set of challenges. At speeds greater than Mach 5, the hypersonic regime, unanticipated aerodynamic heating caused by flow and the effects of friction, sometimes referred to as the “thermal barrier” occurred [10]. Hypersonic aerodynamics was in its early infancy in 1954. The few small hypersonic wind tunnels then in existence had been used almost entirely for fluid mechanics studies. Early wind tunnel development for high-speed flight was pioneered in Germany and later leveraged by the US after WWII. However, many questioned their usefulness for studying hypersonic flight, doubting that the tunnels would be able to accurately simulate real flight conditions (e.g., high heating, temperatures, and Reynolds Number of flight due to strongly interacting flow fields, viscous interactions with strong shocks, and possible real gas effects). In 1954, many believed that new, specialized, true temperature facilities were needed to fully simulate hypersonic flight conditions. In order to break what Dr. John V. Becker called the "facility barrier" [12], all-out efforts were launched to develop high-temperature facilities that accurately simulated hypersonic aerothermodynamic conditions. In the years that followed, NASA’s X-15 program helped to expose the fallacies of the so-called "facility barrier.” Virtually all of the flight pressures and forces observed during X-15 flights agreed with the lowtemperature wind tunnel predictions at Mach numbers10 and below. These findings agreed with analytical predictions that indicated "real gas" high-temperature effects were negligible below Mach 10. Concurrent with the first years of the X-15 flight program, a number of missile and space vehicle configurations were successfully developed in small, low-temperature hypersonic wind tunnels. A few cases obtained limited flight data which provided additional confirmation about the usefulness of low-temperature wind tunnels. With this broad validation, the bulk of which came from the X-15 results, conventional lowtemperature hypersonic wind tunnels became the accepted tool for configuration development (Fig. 6). (Table 1 lists key facilities that studied high-speed and hypersonic flight.)

STO-AVT-234-VKI

PAPER NBR - 3




Figure 6. – Some early NACA hypersonic test facilities [14]. Table 1 – High-Speed/Hypersonic Test Facilities

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned 3.2.1

X-1, X-2 and Supersonics

Breaking the sound barrier—named for the significant challenges (or barriers) that needed to be overcome to achieve supersonic flight—involved solving issues related to high-speed, supersonic aerodynamics addressing flow compressibility effects and shocks, stability and control issues caused by loss of control surface effectiveness, and increased drag. The Bell X-1 and X-2 were the first vehicles that broke the sound barrier, pushing the boundaries of the transonic and supersonic (Mach 1 to Mach 3.2) flight vehicles. These early test vehicles advanced the understanding of high-speed aerodynamics, stability and control, and pilotin-the-loop operability. Advances were also made in several key areas including: liquid rocket engine development and operation; materials and structures to withstand increasing temperatures and heat loads; and inertial coupling, etc. Overcoming flight challenges at low supersonic speeds (Mach 1 to Mach 5) served as the stepping stones that led to the first successful tests at hypersonic speeds with the experimental X-15 airplane. 3.2.2

X-15 – The First Hypersonic Aircraft/Spacecraft

Early ideas for human hypersonic flight were propelled by the advent of large liquid-fuelled rocket engines developed in Germany during WWII. Coupled with NACA’s aeronautics research and DOD support, the X15 program was conceived as a means to understanding and overcoming the thermal barrier. The initial requirements document for the X-15 was surprisingly brief (Table 2). A mere four pages, it was first drafted by NACA LaRC engineers [13] and focused on research, expediency, and using state-of-the-art technology wherever possible.

Table 2 – X-15 Requirements • •

•

• •

Study the effects of flight at Mach 5-7 at altitudes of at least 250,000 ft. (76 km) Study the effects of aerodynamic heating (q = 30 BTU/ft2-s (34W/cm2) • Have representative areas of primary structure experience temperatures of 1,200 F (649 C) Atmospheric Exit and Entry techniques • Reaction control systems • High angle of attack Aerodynamic stability and control in very-low-density air Aeromedical aspects of 0-g

The program quickly advanced. In 1954, NACA completed studies on the challenges of high-altitude, highspeed flight. In December 1955, North American Aviation was awarded a contract to build three X-15 vehicles. The first successful tests of the X-15 vehicle at hypersonic speeds (Mach 5 and above) occurred on June 23, 1961, when US Air Force Major Robert M. White achieved a speed of Mach 5.27. Considered one of the most successful US flight test programs, the 199 flights of the X-15 established a knowledge base that significantly contributed to the successful development of the Space Shuttle Program [13-15]. 3.2.2.1 Aerothermodynamic Heating One of the early, non-intuitive, lessons learned during the progression from supersonic (1 < M 5) was the marked increase in heating rate as a function of radius (q 1/r2). For supersonic flight, streamlined aerodynamics was ideal to minimize drag. For hypersonic flight the opposite was true: leading edges and nose caps with large radii and high angles of attack were needed to minimize the STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned effects of high stagnation heating. This increased-heating effect first came to light during intercontinental ballistic missile (ICBM) tests where the pointed missile nosecones would burn up upon atmospheric entry. In 1951, Alfred Eggers and H. Julian Allen of NASA Ames Research Center (ARC) proposed a “blunt body” configuration for ballistic entry [12] to increase drag, slow the entry vehicle down, and reduce its associated heating. In addition, a blunt body created a strong shock, which was further removed from the vehicle, and reduced the heating to the vehicle even further (Fig. 7). Many early blunt-body shapes were tested in the hypersonic facilities to better understand this phenomenon and to aid in the design, test, and development of hypersonic vehicles.

\

Figure 7. – Research of blunt bodies to reduce stagnation heating [10, 11]. The Allen-Eggers blunt-body concept was incorporated into the X-15’s entry flight profile in the form an increased angle of attack (AOA) during entry (approximately 20 - 26 degrees). Almost 30 years later, the Space Shuttle adopted an optimized AOA during entry of approximately 40 degrees. Another critical lesson learned was the criticality of shock interactions called shock-shock interaction heating. On October 3, 1967, an updated X-15A-2 flew with a pylon-mounted dummy axisymmetric ramjet engine called the Hypersonic Research Engine (HRE) and attained a maximum velocity of Mach 6.70 (6629 fps or 2020 m/s). During the flight, strong shock interactions from the nose of the vehicle, the sharp pylon leading edge, and the static pitot tubes caused severe damage and burn through of the pylon, which caused the engine to detach from the vehicle and crash in the desert below. The lower left portion of Figure 8 illustrates how this behavior corresponds with the prediction of the location that the thin shock-interaction heating layer would impinge on the pylon leading edge. Empirical correlation of elevated temperatures in this region from hypersonic wind tunnel tests revealed that the estimated local increase in aerothermodynamic heating rates was on the order of nine times that of the undisturbed heating to the pylon in that region [16]. Current Computational Fluid Dynamics (CFD) analysis still relies on empirical methods to quantify the magnitude in heating in these regions.

PAPER NBR - 3

STO-AVT-234-VKI




Figure 8 – Effect of Shock-Shock Interaction Heating during X-15A-2 testing of a dummy model of the Hypersonic Research Engine (HRE) [12, 14]. 3.2.2.2 High-temperature structures and materials Operating at hypersonic speeds required significant advances in structural concepts to overcome global (i.e., overall thermal growth and constraint) and local thermal gradients, nonlinear temperature distributions, and stresses. In addition, new classes of metallic and ceramic materials with higher mechanical properties and strengths at elevated temperatures and lower thermal expansion coefficients had to be developed. Maximum temperatures on the X-15 were predicted to be as high as 2,000 °F (1,093 C) requiring the use of a new “superalloy” called Hastelloy-X. In addition, novel concepts for minimizing structural temperatures (hot, insulated and cooled structural concepts) were investigated. For example, the X-15 adopted a hot, heat-sink approach using a thick Hastelloy-X wing leading edge to reduce maximum transient temperatures. It was also slotted; allowing free thermal expansion to relieve associated thermal stresses (Fig. 9). Research revealed that steep, local, nonlinear temperature gradients (Fig. 9) could also cause excessive thermal stresses. Global thermal growth could also lead to catastrophic failure. One example of this is of the X-15 windows which cracked during a flight test due to differential expansion of the window and surrounding metallic frame and structure. Luckily, the canopy glass fracture shown in Figure 9 only damaged the external portion of a two-pane system to contain pressure.

STO-AVT-234-VKI

PAPER NBR - 3




Figure 9. – Global and local thermal stress issues [12]. The X-15A-2 was designed to explore flight at speeds as high as Mach 8. To minimize temperatures, a spray-on silicone ablator called MA-25S was developed by Martin Marietta and applied over the entire vehicle. The total mass of the ablator was approximately 400 lbs. (181 kg) limiting maximum speeds below Mach 7. On October 3, 1967, Pilot Pete Knight flew the first all-ablator-coated version of the X-15A-2 shown in Figure 10 to a record breaking Mach 6.7.

\ Figure 10. – Spray-on-ablator results in protecting aircraft structure [12, 14]. Regions of high heating were evident from the ablator damage on the leading edges, nose cone, and regions with protuberances, shock interferences, and cavities (Fig. 10). The global effects of this heating, which

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned were evident on the ablator-coated X-15A-2, were useful in predicting locations of increased heating and also in empirically calculating quantitative levels of heat flux. It is believed that the challenges with the processing, application, and refurbishment of spray-on ablators for the X-15A-2 influence the decision to use lightweight silica tiles for the thermal protection system (TPS) over acreage areas of the Space Shuttle Orbiter. 3.2.2.3 Hypersonic Aerodynamics and Stability Issues with loss of control and stability experienced during the X-1 and X-2 flights led to a better understanding of thin-airfoil behavior at high-speeds. This knowledge led to the wedge-shaped airfoil sections and the variable split-tail speed brakes on the X-15 horizontal stabilizer and later influenced the shuttle tail design. These modifications helped the X-15 maintain control during its entry AOA of 20 degrees and landing at speeds of over 200 mph (90 m/s) (Fig. 11).

Figure 11. – Hypersonic stability and control [12]. 3.2.2.4 Biomedical and Human Factors The X-15 program also conducted research regarding the human biomedical effects of high-speed, high-g, and sustained low gravity. This led to the development of first practical full pressure suits for pilot protection in space, bio-instrumentation, and flight simulators. Prior to each test flight, flight simulators were extensively used to assess a newly designed stability augmentation system (SAS) as well as a triple redundant adaptive flight control system. 3.2.2.5 X-15 Summary The X-15 program has a long list of accomplishments, which are a direct legacy for Space Shuttle. (See reference 15 and Table 3.) In addition to items discussed in this paper, other leveraged knowledge included: an inertial flight data system for flight both inside and outside the sensible atmosphere; a reaction control system (RCS) that used rockets outside the atmosphere; a blended, adaptive flight control system to seamlessly transition from aerodynamic flight controls to RCS jets; an additional air data system (ADS) during approach and landing; a strategy for bleeding energy during glided Earth entry that later became the terminal area energy management system (TAEM) used for Shuttle; and short nose tricycle-style landing gear.

STO-AVT-234-VKI

PAPER NBR - 3




Table 3 – Knowledge Legacy – X-15 to Space Shuttle [15] • • • • • • • • • • • 3.2.3

Wedge-shaped tail design Blended adaptive flight control system Angle of attack for re-entry and landing Unpowered landings and Terminal Area Energy Management (TAEM) maneuver Extra insulation on leading edges of craft Short-nose tricycle-style landing gear Full pressure suits Advanced inertial flight data system Retractable pitot-tube air data system for landing Fixed ball nose hypersonic flow-direction system Hybrid analogue-digital simulator as well as an overall emphasis on the importance of simulator use.

Lifting Body Research and Tests: Knowledge Transfer to Space Shuttle

In 1957, Dr. Alfred Eggers conceived a totally new class of vehicles he called the “lifting body.” Considered a viable design for manned spaceflight in planes (e.g., X-20) and space capsules (e.g., Mercury, Gemini, and Apollo), the lifting body was a wingless vehicle that flew using lift generated by the shape of its fuselage. Between 1963 and 1975, a total of 222 flights of eight lifting body configurations were tested at NASA’s Dryden Flight Research Center (now Armstrong Flight Research Center) at Edward’s Air Force Base [12, 17, 18]. Several of the lifting body configurations flown are shown in Figure 12, including the X-24-A, M2F3, HL-10, the X-24B, and the X-20 Dynamic Soaring (Dyna-Soar) vehicle. The Dyna-Soar vehicle never flew, however it underwent considerable analysis and technology development. The Air Force’s Dyna-Soar Program, which NASA joined in 1958, sought to develop a reusable; piloted glider with a small payload capability and that could be boosted into orbit atop a Titan II or Titan III. The Dyna-Soar Program accelerated research in hot structures technology such as metallic “shingle”/tile TPS; advanced superalloy materials (Rene-41) and coated refractory-metals; ceramic components and insulation; high temperature bearings, and siliconized graphite (precursor to reinforced carbon-carbon leading edge on shuttle), etc. In addition, advances were made in other areas such as passive and active cooling; correlation of hypersonic entry aerothermodynamics including design detail effects on turbulent heating; and spacesuit design.

. Figure 12. – Early lifting body configurations [17].

PAPER NBR - 3

STO-AVT-234-VKI




4.0 SPACE SHUTTLE DESIGN PHILOSOPHY AND METHODS Beginning with Sänger’s vision of a reusable rocket plane, the successful development and operation of the Space Shuttle was the culmination of many space advocates’ dreams of building a reusable spacecraft to transport crew and cargo to and from space. Studies of the hypersonic flight regime through the methodical analysis, design, and tests of early programs, advanced our understanding of aerodynamic heating and thermal stress; high-temperature materials and insulation; propulsion; and stability and control to such a degree that by the late 1950s and early1960s, both NACA and the DOD were considering the development of reusable space vehicles for both civilian and military uses [18, 19, 20]. There are myriad references on the history of the Space Shuttle Program [19-22] that discuss its complex evolution as it relates to politics, budget challenges, technology barriers, and requirements for the program. Each of these elements played a significant role in the shaping the design and development of the Space Shuttle [21, 22]. George Jeffs, President of Aerospace Operations at Rockwell International, once said, “In concept, the Space Shuttle was to be as close to a ‘state-of-the-art’ system as possible. Developing the system on schedule and achieving the desired reliability left no margin for technology breakthroughs to achieve goals” [20]. The program had to rely on what was learned during previous hypersonic research efforts and space programs (e.g., Mercury, Gemini, Apollo, Skylab, etc.). To address the severe aerothermal environment and requirement for reusability, the size and complexity of the systems and sub-systems, and total mass limits, several technology advances were necessary in several key areas, including: thermal protection systems; a highly efficient ascent propulsion system; improved data processing system (DPS); and structures and materials. A total of six obiters were built, five of which were fully operational. The first orbiter, Enterprise, was designed solely for approach and landing testing. The five vehicles to follow were named Columbia, Challenger, Discovery, Atlantis, and Endeavour. Together, the vehicles flew a combined total of 135 missions from 1981 to 2011. (For a complete summary of all Space Shuttle missions see reference 23.) During its history the program delivered numerous satellites to orbit; captured, retrieved, and repaired satellites and returned payloads back to Earth; launched three interplanetary spacecraft; completed five successful Hubble Space Telescope servicing missions; conducted hundreds of scientific experiments (e.g., life science, combustion physics, Earth observations, etc.); carried more than three million pounds of cargo and over 650 crewmembers into orbit; and served as the primary transportation system during the construction of the International Space Station (ISS). The program suffered two tragedies, Challenger in 1986 and Columbia in 2001, which took the lives of 14 crewmembers. The Space Shuttle satisfied numerous and diverse requirements. Chief among them was to significantly the reduce cost of travel from Earth’s surface to LEO and enable routine access to space. However, the program never achieved this goal. The desire to minimize development costs to meet annual budget limitation shifted the program’s focus away from the recurring costs of operating the shuttle [21]. The refurbishment and repair of major components such as the TPS, Space Shuttle Main Engines (SSMEs), and Solid Rocket Boosters (SRBs) after each mission was costly and time-consuming. The operations costs coupled with the Challenger tragedy in 1986 reduced the projected annual flight rate goal of 50-100 missions to an average annual rate of only 4-5. With large operational costs and so few flights per year, the cost per flight and the price per pound of payload to LEO increased and prevented the shuttle from becoming economically sustainable.

4.1

Space Shuttle Requirements and Concepts

The final requirements for the Space Shuttle were based on many factors, primarily political, budgetary, and technical concerns. In 1970, the support and budget for space decreased, increasing the need for NASA to STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned work closely with the DOD. Their partnership prompted many compromises that were made on the vehicle’s uses and design, which resulted in a broad set of requirements shown in Table 4. The DOD emphasized the need for payload mass and orbit capabilities and a long, crossrange capability to satisfy landings at military airfields (1,100 nm (2037 km)). NASA wanted to significantly reduce costs, provide routine access to space, and build a space station. (The shuttle’s 15-ft. payload bay width is a product of the NASA’s third main requirement, which accommodated space station modules.)

Table 4 – Phase A Space Shuttle Requirements

In 1968 the Space Shuttle Task Group (SSTG) held a meeting to determine the NASA’s needs for space transportation and issued a request for proposal (RFP). This initiated “Phase A” of a multi-phase study for what was then called an Integral Launch and Reentry Vehicle (ILRV) that emphasized economy and safety over the payload performance and requirements specified in Table 4. Over 29 concepts were evaluated [19], ranging from expendable to fully reusable as shown in Figure 13. The SSTG’s final report identified three classes of vehicle concepts for consideration: Class I – reusable orbiter, Class II – reusable orbiter and refurbishable boosters, and Class III – fully reusable (Fig. 14).

Figure 13. – Cartoon of early conceptual design ideas for Space Shuttle.

PAPER NBR - 3

STO-AVT-234-VKI




Figure 14. – Early conceptual design classes studied for Space Shuttle [22]. After the Phase A studies ended in 1970, another RFP called for the initiation of Phase B design studies, which included changes to requirements shown in Table 4. Among them was a specific requirement for the vehicle to be fully reusable and have two stages (Table 5). However in 1971, as vehicle designs matured and predicted staging velocities were calculated to be as high as 12,000 – 14,000 fps, the fully-reusable, twostage requirement imposed greater design challenges by increasing the size of the boosters and the level of heating. Technical problems with the development of TPS and Main Propulsion System (MPS) issues led the designs toward a higher specific impulse engine, later known as the Space Shuttle Main Engine (SSME), and a ceramic fibrous insulation, both of which would require considerable technology development and operations costs. The US government’s decision to only support funding for the Space Shuttle Program (and not a space station program) and limit annual funding were the critical factors in the selection of a Class II stage-and-a-half, partially reusable system. The hope was that this path would minimize yearly development costs to stay within budget limits [19].

STO-AVT-234-VKI

PAPER NBR - 3




Table 5 – Phase B Space Shuttle Requirements

On January 5, 1972 President Nixon announced his approval of the Space Shuttle Program. Later that year, North American Rockwell was awarded the contract to build the Space Shuttle Orbiter, Morton Thiokol was selected to develop the SRBs, Martin Marietta was to develop the External Tank (ET), and Rocketdyne would develop the SSMEs.

4.2

Space Shuttle Design, Operation, and Systems

The final Shuttle design configuration is shown in Figure 15 with primary components and individual elements such as the ground handling systems shown in Figure 16. A list of the primary Shuttle systems is shown in Table 6.

Figure 15. – Space Shuttle final configuration.

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned The two SRBs were selected over liquid fuelled concepts because of their reliability, lower estimated developmental costs, and ease or refurbishment after recovery at sea. They are 149 ft (46 m) long and 12 ft (3.7 m) in diameters and each generated 3.3 million lbs (14.7 million newtons) of thrust. The SRBs were used as a set of matched pairs (i.e., loaded from the same batches of propellant ingredients to minimize thrust imbalance) and were made up of four solid rocket motor segments. The four segments were mated at Kennedy Space Center (KSC) in the Vehicle Assembly Building using an aluminium ring structure and bolts. The ring and bolts that joined motor segments together were known as the field joint. (Problems with the nonlinear deformation of this joint under load and the change in material properties of the O-rings at cold temperatures were determined to be the technical cause of the Challenger tragedy on January 28, 1986.) During launch, six seconds after the SSMEs were ignited and checked out; the SRBs ignited and provided primary steering control for the duration of their 120 seconds of operation. After its fuel was expended, eight booster separation motors were ignited to separate the two boosters from the stack. The expended motors descended under parachutes and were recovered at sea. There, the nozzles were plugged, the water was removed from the casings, and the motor segments were towed back to the launch site and later shipped back to the manufacturer for refurbishment. The SRB nose caps and nozzle extensions were not recovered. The ET was 154 ft (47 m) long and 28.6 ft (8.7 m) in diameter and contained 1.6 million lbs (725,747 kg) of propellant with the liquid oxygen (LOX) tank above the hydrogen tank separated by an intertank region. The liquid hydrogen fuel and LOX oxidizer were supplied under pressure to the three SSMEs during ascent. The ET was attached to the Orbiter at one forward attachment point and two aft attachment points. The liquid oxygen tank was an aluminium monocoque construction in the shape of an ogive which contains 143,351 gallons (543 m3) LOX. The liquid hydrogen was an aluminium semimonocoque construction that contains 385,265 gallons (1,080 m3) of fuel. The ET’s TPS consisted of a spray-on-foam insulation (SOFI) and premolded ablator materials, which maintained cryogenic temperatures of the propellants, eliminated air liquefaction, minimized boil off, and reduced ascent heating to the ET structure.

Figure 16. – Primary Space Shuttle elements (image courtesy and permission of Mr. Bodhan Bejmuk).

STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned The Main Propulsion System (MPS) was composed of the Space Shuttle Main Engines and its controllers, a propellant management system, a helium system, an integrated health management system, hydraulic system, thrust servo actuators, etc. Three new high-performance engine SSMEs, had to be developed to meet the demands of the Shuttle architecture. The engines were reusable and had variable thrust and used liquid hydrogen for fuel and cooling and LOX as the oxidizer. The three SSMEs were configured in a triangular arrangement and each had to perform at a thrust level of 375,000 pounds-force (lbf) at sea level and 470,000 lbf in a vacuum (corresponding to a rated power of 100%), and 417,300 lbf at sea level and 513,250 lbf in a vacuum (corresponding to 109%). The SSMEs burned 750 gallons (2.8 m3) of hydrogen and 280 gallons (1.1 m 3) of oxygen per second. The Space Shuttle Orbiter was constructed primarily of aluminium for load bearing structures and was protected externally by reusable surface insulation (RSI) TPS. The highly heated sections like the nosecap and wing leading edges (WLEs) were fabricated from reinforced carbon-carbon (RCC), which could withstand temperatures and maintain structural properties up to 3,000 °F (1649 C). It was divided into nine major sections: 1) the forward fuselage, 2) wings, 3) mid-fuselage, 4) payload bay doors, 5) aft fuselage, 6) forward reaction control system (RCS), 7) vertical tail, 8) Orbital Maneuvering System (OMS)/RCS pods, and 9) body flap. The Orbiter is 122 ft (37 m) long and 57 ft (17 m) high and has two delta wings with a span of 78 ft (24 m). The Orbiter body flap thermally shielded the three SSMSs during Earth entry and provided pitch control trim during atmospheric flight after entry. The vertical tail consisted of a structural fin surface, a rudder/speed brake surface, a tip, and a lower trailing edge. The rudder split into two halves to serve as a speed brake similar to the vertical tail of the X-15 (Fig. 11). A complete list of primary Space Shuttle Systems is shown in Table 6, and described in detail in references 20 and 24. The remainder of the paper will highlight some of the lessons learned that can be related to the themes mentioned earlier in this paper.

Table 6 – Space Shuttle Systems

4.2.1

Space Shuttle Operation

Space Shuttle launched vertically; the SRBs and SSMEs burned in parallel with 6 a second delay in SRB ignition to allow checkout of the SSMEs to rated power (usually 104%) (Fig.17). At a velocity greater than or equal to 127 fps (39 m/s) the vehicle rolls, pitches, and yaws to a heads-down ascent attitude. Maximum dynamic pressure (Qbar) was reached at approximately Mach 2.2 (about 30 - 60 seconds after liftoff), and at that time the SSMEs were throttled down to remain within structural limits and then raised back up to 104

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned percent. Each SRB used hydraulic servo-actuators, which were powered by self-contained independent hydraulic power units (HPUs) fuelled by hydrazine to gimbal the nozzles for thrust vector control (TVC) during the ascent phase. Approximately two minutes into the ascent phase, the two SRBs consumed their propellant, chamber pressures dropped below 50 psi (3.5 kg/cm2) and they were jettisoned from the ET following a separation signal sent from the Orbiter. At approximately 188 seconds after separation at an altitude of 15,700 ft (4,785 m), pilot drogue chutes were deployed to stabilize the SRBs in a tail-first configuration. At 5,500 ft (1,676 m) three main chutes were deployed and the SRBs splashed down in the ocean and awaited recovery, ferry to port, and subsequent refurbishment. Following SRB separation, the Orbiter and ET continued to ascend while small thrusters carried the SRBs away from the “stack.” At about 8 minutes and 30 seconds after launch, the three SSMEs underwent a controlled main engine cutoff (MECO). The ET was jettisoned on command from the Orbiter and forward and aft reaction control system (RCS) jets were fired to provide attitude control of the Orbiter. This maneuvered it away from the ET and positioned the Orbiter in the appropriate burn attitude prior to the orbital maneuvering system (OMS) burn for orbit insertion. The ET burns up during Earth entry and splashes down in the ocean.

Figure 17. – Space Shuttle mission operations. The first OMS burn (OMS 1) placed the Orbiter in a stable elliptical orbit referred to as “direct insertion.” (Altitudes varied from 100 to 312 nm (185 to 578 km)). The crew then performed a second OMS burn (OMS 2) to circularize the orbit. On orbit, the RCS jets provided attitude control for the Orbiter as well as any minor maneuvers while on orbit. The OMS engines were used to perform orbital maneuvers such as those to rendezvous with the ISS. While on orbit the Shuttle crew (usually 7 crewmembers: 1 commander, 1 pilot and 5 mission specialists) performed mission objectives such as: logistics (equipment, supplies, experiment rack transfer, etc.), payload deployment or retrieval, ISS assembly, and performing scientific experiments. At the end of orbital operations, the RCS jets oriented the Orbiter in a tail-first attitude in preparation for an

STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned OMS burn, which decreased the Orbiter’s orbital velocity anywhere from 200 to 500 fps (61 to 152 m/s) in preparation for Earth entry. After the OMS burn completed, the RCS jets rotated the vehicle to a nose-first attitude. Entry interface (EI), the altitude where the vehicle is considered to enter the atmosphere, was considered to occur at 400,000 ft (122 km). At this point, the vehicle was approximately 4,200 nm (7,778 km) from the landing site (nominally KSC) and the Orbiter maneuvered to 0-degree roll and yaw and a 40degree AOA for entry. Entry guidance dissipated the tremendous energy from orbit through heat as the vehicle slowed through Earth’s atmosphere. It controled entry angle to prevent skipping out of the atmosphere, while the AOA and bank angles to managed heating for the Orbiter structure and TPS. The entry terminal area energy management (TAEM) interface was located at an altitude of 83,000 ft (25 km) and at an AOA of approximately 14-degrees. At a velocity of 2,500 fps (762 m/s) and a range of 60 nm (111 km) from the runway, TAEM guidance commanded the vehicle to manage energy and direct the vehicle to one of two heading alignment cones (HAC) about the runway. The vehicle touched down between 195 to 205 knots (100 to 105 m/s). Typically, the Orbiter landed at KSC. During emergencies or hazardous weather, the vehicle could land on the dry lakebed at Edwards Air Force Base in California. There were several alternate landing sites in France and Spain in the event of an abort situation during ascent, which was known as transAtlantic landing (TAL).

Figure 18. – Kennedy Space Center (KSC) ground turnaround sequence [24]. 4.2.2

Space Shuttle Turnaround

Upon landing, a ground team immediately approached the vehicle to ensure it was safe, there were no leaks, and the surrounding environment was safe. An air conditioning purge unit was attached to the Orbiter’s aft end so air could be circulated to cool the aft end, wings payload bay, forward fuselage, and OMS pods to dissipate the heat of entry (Fig. 18). A second ground cooling unit was connected to the vehicle to cool the flight crew and avionics after landing. Once the flight crew egressed the vehicle, the Orbiter was powered down, and the vehicle TPS and systems were thoroughly checked prior to towing to the Orbiter Processing PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned Facility (OPF). (The post-landing procedures were the same had a vehicle landed at Edwards Air Force Base. The vehicle would have been transported back to KSC atop the Shuttle Carrier Aircraft.) Figure 19 compares the notional view of aircraft-like operations of the Orbiter, which was a much desired objective, to an actual view from the OPF facility during a real Shuttle turnaround. As mentioned earlier, miscalculations in the technology readiness and robustness of key shuttle components such as the TPS, SSME, and SRBs significantly lengthened the time required for inspections, repairs, replacements, and refurbishments. The Space Shuttle was a thoroughbred, not the workhorse originally envisioned. It required intense pampering between missions. The shuttle cost approximately $4 billion per year and took 18,000 people on the ground to operate it. To achieve low-cost, routine access to space will require a vehicle with a robust propulsion system and damage-tolerant TPS that has been validated through rigorous testing to ensure ease-of-operation (e.g., installation, inspection, servicing, etc.) to withstand the severe environment of repeated launch, orbit, Earth entry, and landing.

Figure 19. – Space Shuttle operations goal vs. operations reality (image courtesy and permission of Mr. Bodhan Bejmuk).

4.3

Loads

Figure 20 illustrates some of the many load cases that needed to be considered during the certification process to ensure that all constraints were satisfied prior to launch. Early in the Space Shuttle Program, simple methods related to each discipline for predicting and determining aerodynamic, aerothermodynamic, static and dynamic structural loads and performance were used. Examples included:    

Simple beam and plate methods for estimating stresses and thermal/structural loads for initial structural sizing to evaluate overall system configuration trade studies and designs. Finite-element structural models to determine loads, deflections, and stresses, and stepwise buildingblock tests. Aerothermal heating rates to simple shapes (e.g., cones, cylinders, spheres and plates) estimated stagnation heating and heat-flux distributions. Computational fluid dynamics (CFD) methods were later used to calculate detailed flowfield behavior as well as aerothermal heat loads.

Many facilities were constructed for simulating low-speed, supersonic, and hypersonic flows; entry heating conditions; static, dynamic, and acoustic loading; etc., (see Figure 21) for testing critical shuttle components. Local analytical methods were used to size the TPS and verified by radiant and aerothermal tests in wind tunnels and arcjets. The science of wind tunnel testing had matured since The 1950’s and large facilities such as the 8-Foot High Temperature Tunnel at NASA Langley Research Center (LaRC) were used to

STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned conduct tests on models large enough to test and observe flow and aerothermal heating to complex arrays of shuttle tiles attached to real structural models (Fig. 21). Dynamic and acoustic test facilities were also developed to test large-scale models of shuttle components. As shown in Figure 20, three launch-to-orbit configurations were considered to assess loads, integration, and operational issues: 1) shuttle on the ground and on liftoff, 2) post-liftoff configuration, and 3) “boost” configuration. Liftoff loads, ground winds, acoustic loads, tank pressurization levels, thermal structural loads on tanks were critical during the first phase. Minor details such as mobile launch platform (MLP) compliance could have a significant impact on induced dynamic loads to the vehicle. Winds aloft, aerodynamic loads during times of high dynamic pressure (max q), aeroheating and plume effects, flutter & buffet, acoustic loads, POGO, etc., were considered during liftoff.

Figure 20. – Shuttle Load configurations and cases (liftoff to orbit). (Image courtesy and permission of Mr. Bodhan Bejmuk).

PAPER NBR - 3

STO-AVT-234-VKI




Figure 21. – Some of the many models and test facilities utilized to conduct configuration trade studies and develop and ensure safe operation of Space Shuttle (images courtesy of Mr. Dennis R. Jenkins). 4.3.1

Day of Launch Initialization Load (I-Load) Update (DOLILU)

In order to minimize loads during launch, maximize mass to orbit, and ensure safety of flight, the Space Shuttle Program (SSP) optimized the vehicle’s steering commands with day-of-launch environmental conditions (e.g., ground winds and winds aloft, wind direction, and atmospheric thermodynamic data). The DOLILU process is illustrated in figure 22 [26]. Scheduled balloon launches on the day of launch provided high-resolution wind speed and direction up to 58,000 ft (17.7 km) and a forecasted wind from 0 to 80,000 ft (24.4 km) to identify large wind variations. Wind data was used to help design the vehicle trajectory (angleof-attack (Alpha ()), angle-of-sideslip (Beta ()), and maximum dynamic pressure (Qbar)) to ensure structural loads were within limits at points during the trajectory. Qbar, occurs around Mach 2.2 at an approximate altitude of 58,000 ft (17.7 km) and 1,500 ft (.46 km), during the single axis rotation (SAR) or roll maneuver. Structural constraints during launch were predicted using a simplified set of structural load indicators (SLIs), which are a complex set of algorithms based on a more rigorous set of stress analyses. The wind persistence and dispersions were used to calculate the possible error and define the limit boundaries and margin. The points of the trajectory together with an associated dispersion ellipse are plotted on a “QPlane” as a function of Alpha and Beta for every Mach number and, if the Qbar point is determined to be within limits, there is a “go for launch.” The launch process also includes an Independent Verification and Validation (IV&V) Team with a separate chain of command that independently receives wind data and verifies recommendations of the DOLILU team.

STO-AVT-234-VKI

PAPER NBR - 3




Figure 22. – Day of Launch Initialization-Loads (I-Loads) Update (DOLILU) Overview [26]. The DOLILU process evolved over ten years from a single I-Load method used on earlier flights to the process described above [25]. (Results are shown in Figure 23.) Launch probability predictions were improved from 50% (i.e., more than half of the measured winds aloft violated the vehicle’s certified boundaries) to less than 5%. Hence, a lesson learned was that although the operations overhead was expensive, if you commit to such a process during early development, you can significantly improve margin and launch probability.

Figure 23. – Day of Launch Initialization-Loads (I-Loads) Update (DOLILU) evolution (Image courtesy and permission of Mr. Bodhan Bejmuk).

PAPER NBR - 3

STO-AVT-234-VKI




4.4

Structures and Materials

Instead of the hot structure, heat-sink approach used for the X-15, the Space Shuttle Orbiter used an aluminium primary structure with external insulation (the lightest configuration) to manage the high temperatures of hypersonic flight [19, 22]. As mentioned earlier, material development was necessary to mature several new technologies for TPS for the entire vehicle. Reusable Surface Insulation (RSI) was developed to accommodate different temperature regions and had specific designations such as:  High-Temperature Reusable Surface Insulation (HRSI) tiles, which were fibrous ceramic tiles on the belly of the vehicle, capable of withstanding temperatures from 1200 to 2300 °F (649 to 1260 C);  Low-Temperature RSI (LRSI), which were white tiles on the sides of the vehicle and the OMS pods where temperatures were moderate (temperatures from 700 to 1200 °F (371 to 649 C));  Flexible RSI (FRSI), which were coated Nomex felt blankets that covered the upper, lower heated regions of the vehicle as shown in Figure 24. There are approximately 25,000 to 30,000 tiles and 3,000 flexible blankets covering the vehicle’s outer surface. The nosecap and wing leading edge were made of Reinforced Carbon-Carbon (RCC) capable of withstanding temperatures as high as 3000 °F (1649 C). There were 22 individual RCC panels, which were separated by T-seals made of RCC to allow differential expansion and prevent hot-gas ingress during entry. The Shuttle tile material was very low-density, fibrous ceramic material (~90% air), which had a low coefficient of thermal expansion (CTE) and was very brittle (i.e., low strain to failure). To reduce global and local mechanical and thermal loads and stress, the size of the tiles was restricted to 3 x 3 in (7.6 x 7.6 cm). The RCC wing leading edges were segmented (similar to the X-15 leading edges) into 22 sections each to minimize global stresses resulting from the wing bending and thermal stresses caused by nonlinear temperature gradients, and relieve thermal stress by allowing differential thermal expansion (as noted earlier, a segmented leading edge was selected for the X-15 for similar reasons). Serious issues with local stresses caused technical problems during the development of both the RSI tiles and the RCC leading edges. Both systems required significant development time and an inordinate amount of operations time for inspection, repair, and/or refurbishment after every flight. This miscalculation in operations costs prevented the Space Shuttle from realizing its goal of inexpensive, routine access to space.

Figure 24. – Space Shuttle thermal protection system (TPS) map [31]. 4.4.1

The Shuttle Tile Story – “The Devil is in the Details”

Prior to the launch of Space Transportation System-1 (STS-1) on April 12, 1981, after flight profiles and air

STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned loads were refined, it became apparent that while the individual components that make up the RSI satisfied loading requirements, the RSI as a “system” lost about 50% of its original tensile strength. This meant that many of the shuttle tiles would exceed structural limits and fail. At the time, Rockwell had already installed over 24,000 tiles on the vehicle before the root cause of problem was finally discovered and a solution was in hand. On the shuttle ferry flight from Palmdale, California to KSC, a large number of tiles fell off. Loss of even one HRSI tile on the bottom of the orbiter in a critical location could cause burn through of the aluminium structure, hot gas ingress, and loss of the vehicle during the high heating phase of entry, which lasts approximately 15 minutes. NASA created a “Tiger Team” led by Dr. Paul Cooper at LaRC, which included scientists and engineers from multiple NASA Centers to determine the cause of this very serious problem and recommend a solution [27]. The Space Shuttle’s TPS experienced a variety of loads during ascent, which had to be considered in determining limit loads to the system: 1) SSME oscillatory pressure waves and acoustic loads (as high as 165 dB) could excite the aluminium substructure and cause TPS inertial loads; 2) ascent aerodynamic loads: aerodynamic pressure gradients, acoustic pressure loads caused by boundary layer (BL) noise, shocks, buffet, gust, and unsteady loads from vortex shedding from adjacent shuttle elements; 3) differential pressures from outgassing of the porous tiles during launch as tiles and the strain isolator pad (SIP) vent; 4) flight loads that cause out-of-plane deformation of the substructure; 5) thermal stresses caused by CTE mismatch between the tile, the reaction cured glass (RCG) coating, aluminium substructure, and SIP; and 6) entry loads from out-of-plane substructure deformation, BL noise, BL separation, maneuver and landing loads, etc. The shuttle tile problem was so difficult to solve; it delayed the first launch of Space Shuttle (STS-1) by almost one year. During that time, the NASA Tiger Team conducted an extensive research and test program, which included material property testing of each of the tile components as well as the system of components; tension and compression tests, shear tests, combined static tension and moment tests, cyclic fatigue tests, photoelastic tests, and non-destructive tests. RSI analyses included static, dynamic, nonlinear dynamic, material nonlinear analysis, fatigue analyses, etc. [27 to 29].

Figure 25. – Space Shuttle thermal protection system (TPS) tile assemblage [22, 31]. To alleviate loads in the brittle ceramic tiles, in particular thermal stresses due to thermal expansion To PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned alleviate loads in the brittle ceramic tiles, in particular thermal stresses due to thermal expansion mismatch between the tiles and aluminium substructure, a flexible strain isolator pad (SIP) was bonded between the tile and the aluminium. The SIP is a Nomex felt pad/mat of fibers held together by pulling a barbed needle repeatedly through the transverse direction during manufacture. This process causes the material to be heterogeneous with local, discreet fiber bundles through the transverse direction as shown in Figure 26. The SIP is coated on both surfaces with a room-temperature-vulcanizing (RTV) silicone adhesive that was used to bond the tile to the substructure. During the course of testing, it was discovered that a failure at the TileSIP bondline, a relatively small, microscopic detail, was the root cause of the tiles falling off (Fig. 26). The increased localized stiffness caused by the fiber bundles resulted in local stress concentrations that reduced the flatwise-tensile strength of the TPS system by approximately 50%, causing premature tile failures. The ability to identify the root cause of premature tile failures enabled the team to develop a simple solution: densify the bonding surface of the RSI tiles with ceramic slurry [27]. This densified layer was 0.02 in (.05 cm) thick and provided a hard, strong, nearly continuous layer that distributed the load and strengthened the tile-SIP interface. This increased the tensile strength and allowed failures to occur in the parent, undensified tile material.

Figure 26. – Photoelastic test results indicate increased local stress concentrations caused by stiff, transverse fiber bundles in the strain isolator pad (SIP) as the cause of premature Shuttle tile failures [30, 31]. 4.4.2

Reinforced Carbon-Carbon (RCC) Material and Wing Leading Edges

Reinforced Carbon-Carbon (RCC) is not a material, per-se, as much as it is a “material system” or “system of materials.” As mentioned earlier, RCC was developed for use in ballistic missile nosecones, however, unless it has a sufficient oxidation protection, it will ablate during atmospheric entry. Hence, developing and certifying this material for extended reuse was a development challenge for the Shuttle Program. RCC is a composite made by curing graphite fabric that has been pre-impregnated with phenolic resin and laid up in molds for a desired shape [32]. The polymeric resin is converted to carbon by pyrolysis—a chemical change brought about by the addition of heat. The part is then densified to increase its mechanical properties by STO-AVT-234-VKI

PAPER NBR - 3



Space Shuttle Design and Lessons Learned repeated cycles of infiltrations with furfuryl alcohol and heating. A thin, silicon carbide (SiC) oxidation protection outer layer for the carbon-carbon (C-C) is formed by a diffusion coating process that converts the outer three plys on the C-C part to form SiC [Fig. 27]. The thin (0.04 in (0.1 cm)) SiC layer is a different CTE than the underlying C-C substrate and, hence, thin cracks form through the SiC layer during processing cooldown as shown on the right in Figure 28 [33, 34]. The thin cracks could serve as pathways for oxygen to penetrate the subsurface material to oxidize it and cause voids [33, 34] (left side of Fig. 28). These voids beneath the surface of the RCC are critical because they could jeopardize the integrity of the RCC coating, cause oxidation, burnthrough of the RCC panels, and hot gas ingress during Earth entry, which could result in a catastrophe. To help prevent this problem, tetraethyl orthosilicate (TEOS) is applied via vacuum impregnation to fill remaining porosity. A Type-A sodium silicate glass sealant is then applied to the outer surface to fill craze cracks and impede oxygen penetration to the C-C substrate. The Type-A sealant fills the craze cracks and after curing, forms a glass outer coating. The craze cracks closed during entry heating and since the glass is viscous at those temperatures, some of the glass will flow onto the outer surface and be driven away by shear flow forces. Hence, because of the unknown reusability of this material, a rigorous inspection test program was instituted to ensure subsurface mass losses and potential coating integrity issues would not cause a catastrophic failure. However, the devil was in the details, and selection of a relatively new technology without sufficient analysis and testing to verify its performance led to a standing army of people on the ground and caused operations costs to skyrocket. Also, this material was not understood well enough as a system (thermal, structural, material, etc.) to identify and predict key critical failure mechanisms prior to design selection and use.

Figure 27. – Reinforced Carbon-Carbon (RCC) material system [32].

PAPER NBR - 3

STO-AVT-234-VKI




Figure 28. – Reinforced Carbon-Carbon (RCC) material craze cracks and voids [33, 34]. As mentioned earlier, the wing leading edges were composed of 22 individual RCC panels and RCC Tseals to allow overall wing mechanical growth under load and thermal expansion differences between panels and attachment structures (Fig. 29). Because the leading edges operate as a hot structure and could reach temperatures as high as 3000 °F (1649 C) during entry, the leading edge could radiate a portion of that heat to space to cool the hotter regions such as the stagnation region. In addition, internal radiation from the lower (hotter) wing surface could radiate to the upper (cooler) surface to help cool and equalize temperatures. Each RCC panel and T-seal had a unique shape and; because of strict fit tolerances to prevent hot gas ingress and also allow differential thermal expansion; each panel component is custom fit to each vehicle and wing segment (port and starboard). The forward spar was covered with insulation and a thin superalloy shield to protect it from high temperatures within the wing cavity and potential sneak flow of hot-gas ingress during entry. Installation of a Shuttle RCC wing panel at KSC to the forward spar is shown in the lower left of Figure 29. Throughout the 30-plus years of the Shuttle Program there were many issues concerning leading edges that were addressed. Rigorous procedures were developed to inspect the leading edges using visual means, tactile methods, non-destructive evaluation (NDE) (e.g., ultrasound (C-scans), X-ray, infrared camera, etc.), and destructively by using plugs to monitor the rate of subsurface mass loss and void growth. Until the loss of Columbia during the STS-107 mission on February 1, 2003, many of the engineers had believed the RCC wing leading edges to be durable, robust, and capable of surviving a large foam debris strike from the ET during launch. After the Columbia tragedy, the key technology working group responsible for the RCC wing leading edges, the Leading Edge Structural Sub-System Problem Resolution Team (LESSPRT), and the nation learned how much they did not know about the reliability, impact damage tolerance, and entry survivability of the Shuttle. Furthermore, the tragedy revealed knowledge gaps about the causes of systemic cyclic exposure damage to RCC leading edges. Even the analytical methods used to assess impact damage to the wing while the STS-107 crew was on orbit were found to be inadequate [4].

STO-AVT-234-VKI

PAPER NBR - 3




Figure 29. – Reinforced Carbon-Carbon (RCC) wing leading edge system. 4.4.3

External Tank (ET) Thermal Protection System (TPS)

Although the ET was not reusable, its TPS plays a vital role in the performance and safe operation of the Space Shuttle prior to and during launch. The ET contained the cryogenic propellants liquid oxygen (LOX) and liquid hydrogen, and delivered them through feed lines to the three SSMEs. In addition, the ET provided the structural backbone of the Space Shuttle stack, connecting the two SRBs and Space Shuttle Orbiter. The ET TPS is composed of many forms of insulation: closed-cell spray-on foam insulation (SOFI) throughout most of its exterior, hand-applied insulation for certain regions, and ablative protection in certain areas (Fig. 30). The TPS maintained the temperature of the cryogenic propellants: -423 °F (-253 C) for the liquid hydrogen and -296 °F (-182 C) for the LOX. It also minimized propellant boiloff, protected the structure during ascent heating, prevented ice formation prior to launch (a potential debris threat to the Orbiter and its fragile TPS), and prevented the liquefaction of air to any exposed aluminium tank surface.

PAPER NBR - 3

STO-AVT-234-VKI




Figure 30. – External Tank (ET) Thermal Protection System (TPS) [31]. The SOFI was applied using a controlled, automated procedure whereby the SOFI thickened using layer after of layer of over-spray applications. This process caused “Knit” lines to form as previous layers cured during the automatic processing. The Knit lines were thin, stiff layers with different mechanical properties than the acreage foam (see Fig. 31). In addition, the material was highly nonlinear, anisotropic, and heterogeneous, making complete material property characterization, detailed stress analysis, and failure mode predictions virtually impossible. In addition, even though the foam was closed-cell, moisture could diffuse into it over time and potentially cause local failures during launch through Earth’s atmosphere and into the vacuum of space. Hence there were many different failure mechanisms for the ET TPS depending on the location on the cryotank, the specific processing method, loading, etc.

STO-AVT-234-VKI

PAPER NBR - 3




Figure 31. – External Tank (ET) Thermal Protection System (TPS) [4]. In addition, even though the foam is closed cell, moisture can diffuse into the foam over time and potentially cause local failures during launch through the Earth’s atmosphere and into the vacuum of space. The foam system had approximately 12 defects which have been identified [35] (shown in Fig. 32) and which could be the initiation sites for local failures which had to be assessed.

Figure 32. – External Tank (ET) Thermal Protection System (TPS) foam defect assessment (ref. 35).

PAPER NBR - 3

STO-AVT-234-VKI



Space Shuttle Design and Lessons Learned There were many different failure mechanisms for the ET TPS depending on the location of the cryotank, the specific processing method (automatic/robotic spray, hand layup, combination of different materials and system layups (e.g., various foams (BX-265, NCFI 24-57, and PDL-1034) and ablator materials (SLA-561 and MA-25S (similar to the ablator used on the X-15A-2)) [31].

Figure 33. – External Tank (ET) substrate debonding failure mode [36]. Figure 33 illustrates only one type of failure mode, substrate debonding, which opens a through-thethickness crack which could allow air to liquefy upon reaching the aluminium tank surface and continue to be pumped into the crack by a process called cryopumping! ET foam debris was never intended to shed from the cryotank surface and cause damage to the Shuttle Orbiter TPS. It was never intended that such damage could cause a catastrophic failure as it did on STS-107; however, ET foam debris shedding occurred from the very first Shuttle launch and continued till its retirement. An example of the types and location of foam loss during STS-114 in 2005 is shown in Figure 33 [37]. Very minor changes to even the blowing agents in the foam caused a significant increase in debris and damage to the Orbiter tiles. Once again, design of a sensitive as opposed to a robust vehicle caused unanticipated damage and costly refurbishment and loss of life! Because of the difficulty in understanding the root cause of the bipod foam failures, the SSP decided to redesign the fitting and eliminated the bipod foam ramp after it was determined it would not cause undue aerothermal or thermal-structural consequences to the vehicle (Fig. 33).

STO-AVT-234-VKI

PAPER NBR - 3




Figure 33. – External Tank (ET) foam debris loss areas following the launch of STS-114 [37].

5.0 CONCLUDING REMARKS The Space Shuttle Program’s storied history is vast and well documented. Understanding the design and operations of this unique and complex vehicle is not confined to the study of one program, but of many. It touches on only a handful of the lessons that were learned through the various supersonic and hypersonic research programs that laid the foundation for Shuttle. In addition, a select handful of recurring themes which transcend the various research programs are discussed. The engineers and leaders charged with the development of next-generation exploration vehicles should take care to integrate the knowledge gained from these programs into future designs. As Petroski once wrote, “The design is us.” [34] Vehicles and programs are the product of human creativity and ingenuity. Consequently, they can also reflect our biases, misunderstandings, errors, and oversight. Continuous learning from both failure (i.e., what was done poorly) and success (i.e., what was done well) is critical. Through rigorous and robust processes and systems thinking, the hope is that preventable failures will be prevented and that our designs will be resilient enough to adapt to the unexpected.

6.0 ACKNOWLEDGEMENTS I would like to acknowledge the careful editorial assistance of Ms Haley Stephenson, the technical discussions and assistance of Gen. Joseph H. Engle, and the historical discussions of Michael Ciancione and Dennis R. Jenkins. I would also like to thank Mr Bohdan Bejmuk and Mr Jenkins for use of their figures.

PAPER NBR - 3

STO-AVT-234-VKI




7.0 REFERENCES 1. Camarda, Charles J.; Bilen, Sven; de Weck, Olivier, Yen, Jeannette; and Matson, Jack: “Innovative Conceptual Engineering Design – A Template to Teach Problem Solving of Complex Multidisciplinary Design Problems.” American Society for Engineering Education Annual Exposition and Conference, Louisville, Kentucky 2010. 2. Camarda, Charles J.: “Failure is not an Option…It’s a Requirement.” AIAA Paper Number 2009-2255. Presented at the 50th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, Palm Springs California, May 2009. 3. Johnson, Theodore F.; Natividad, Roderick; Rivers, H. Kevin; and Smith, Fussell W.: “Thermal Structures Technology Development for Reusable Cryogenic Propellant Tanks.” NASA TM-2005213913, September, 2005. 4. Gehman, H. W., et. al,: “ Columbia Accident Investigation Board.” Report Volume 1, U. S. Government Printing Office, Washington D. C., August 2003. http://www.nasa.gov/columbia/home/CAIB_Vol1.html 5. Roberto, Michael A.: “Lessons from Everest – The Interaction of Cognitive Bias, Psychological Safety, and System Complexity.” California Management Review, Vol. 45. No. 1, Fall 2002. 6. Petrowski, Henry: “To Engineer is Human – The Role of Failure in Successful Design.” Vintage Books, 1992. 7. Petrowski, Henry: “Design Paradigms – Case Histories of Error and Judgment in Engineering.” Cambridge University Press, 1994. 8. Matson, Jack V.: “Innovate of Die – A Personal Perspective on the Art of Innovation.” Paradigm Press Ltd., 1996. 9. Edmundson, Amy C.: “Strategies for Learning from Failure.” Harvard Business Review, The Failure Issue, April, 2011. 10. Heppenheimer, T. A.: “Facing the Heat Barrier – A History of Hypersonics. NASA SP 2007-4232 Sept. 2007. 11. Hallion, Richard P. “The Hypersonic Revolution: Eight Case Studies in the History of Hypersonic Technology - Volume I, From Max Valier to Project Prime (Aeronautical Systems Division, WrightPatterson AFB, Ohio, 1987). 12. Becker, John V.: The X-15 Program in Retrospect.” Paper presented at the 3rd Eugen Sanger Memorial Lecture. Bonn, Germany, 4-5 December 1968. http://www.hq.nasa.gov/office/pao/History/x15lect/hyper.html . 13. Jenkins, Dennis R.: “Hypersonics Before the Shuttle – A Concise History of the X-15 Research Airplane.” Monographs in Aerospace History Number 18. NASA SP-2000-4518, June 2000.

STO-AVT-234-VKI

PAPER NBR - 3




14. Jenkins, Dennis R.: “X-15 – Extending the Frontiers of Flight.” NASA SP-2007-562, 2007. 15. Anon: “Knowledge Legacy – From the X-15 to the Space Shuttle.” NASA Academy of Program/Project and Engineering Leadership (APPEL). http://appel.nasa.gov/knowledgesharing/case-studies/appel-case-studies/knowledge-legacy-from-the-x-15-to-the-space-shuttle/ 16. Watts, Joe d.: “Flight Experience with Shock Impingement and Interference Heating on the X-15-2 Research Airplane. NASA TM X-1669. June 1968. 17. Reed, R. Dale: “Wingless Flight – The Lifting Body Story.” NASA SP-4220, 1997. 18. Wallace, Lane E.: “Flights of Discovery – 50 Years at the NASA Dryden Flight Research Center. The NASA History Series, NASA SP-4309, 1996. 19. Williamson, Ray A.: “Developing the Space Shuttle.” Exploring the Unknown – Selected Documents in the History of the U.S. Civil Space Program. Chapter Two, Volume IV: Accessing Space. SP 4407, 1999. 20. Jeffs, George W.: “The Space Shuttle Design and Construction.” Rockwell International, March 1979. 21. Logsdon, John M.: “The Decision to Develop the Space Shuttle.” Space Policy, May 1986. 22. Jenkins, Dennis R.: “Space Shuttle – The History of the National Space Transportation System: The First 100 Missions.” 2010 Dennis R. Jenkins Publisher. 23. Legler, Robert D. and Bennet, Floyd V.: “Space shuttle Mission Summary.” NASA TM-2011216142, September 2011. 24. Anon: “Shuttle Crew Operations Manual.” NASA Contract NAS9-20000, United Space Alliance (USA), 2004. 25. Bejmuk, Bo: “Space Shuttle Integration Lessons Learned.” Presentation to NASA JSC, December 2005. 26. Harrington, Brian E.: “Space Shuttle Day-of-Launch Trajectory Design Operations. Conference Paper, Annual Technical Symposium, Houston, TX April 20, 2010

AIAA

27. Cooper, Paul A. and Holloway, Paul F.: “The Shuttle Tile Story.” Aeronautics and Astronautics, Vol. 19, No. 1, Jan. 1981, pp. 24-34, 36. 28. Sawyer, James Wayne: “Mechanical Properties of the Shuttle Orbiter Thermal Protection System Strain Isolator Pad.” Journal of Spacecraft, Vol. 21, No. 3, May-June 1984.

PAPER NBR - 3

STO-AVT-234-VKI




29. Ransone, Philip O. and Rummler, Donald R.: “Microstructural Characterization of the HRSI Thermal Protection System for Space Shuttle.” NASA TM 81821 May 1980. 30. Cooper, Paul A. and Prabhakaran, R.: “Photoelastic Tests on Models of Thermal Protection System for Space Shuttle.” NASA TM 81866, August 1980. 31. Hale, Wayne; Lane, Helen; Chapline, Gail; and Lulla, Kamlesh: “Wings in Orbit – Scientific and Engineering Legacies of the Space Shuttle, 1971-2010.” NASA SP-2010-3409, 2010. 32. Gordon, Michael P: “Leading Edge Structural Subsystem and Reinforced Carbon-Carbon Reference Manual”. Boeing Report KL0-98-008, October 19, 1998. 33. Jacobson, Nathan S.; Roth, Don J.; Rausser, Richard W.; and Curry, Donald M.: “Oxidation Through Coating Cracks of SiC-Protected Carbon/Carbon, NASA TM 214834, June 2007. 34. Presentation of the status of OV-105 nosecap damage by Mike Gordon (subsystem manager (SSM) of the LESS) on 4/22/04 entitled: “OV-105 Nose Cap Damage”. 35. Hilburger, Mark W. and Nemeth, Michael P.: “Application of Video Correlation Techniques to the Space Shuttle Tank Foam Materials.” AIAA Paper Number 2006-2119-311 presented at the 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, May 1-4 2006, Newport Rhode Island. 36. Weiser, E. S.; Nemeth, M. P.; and St. Clair, T. L.: “Assessment of Technologies for the Space Shuttle External Tank Thermal Protection System and Recommendations for Technology Improvement, Part 1: Materials Characterization and Analysis.” NASA TM-2004-213238, July 2004. 37. Gomez, Reynaldo J. III: “STS-114 Foam Debris Potential Damage.” NASA JSC, December 9, 2005 38. Knight Jr., Norman F.; Nemeth, M. P.; and Hilburger, Mark W.: “Assessment of Technologies for the Space Shuttle External Tank Thermal Protection System and Recommendations for Technology Improvement, Part 2: Structural analysis Technologies and Modelling Practices.” NASA TM-2004-213256, July 2004.

STO-AVT-234-VKI

PAPER NBR - 3
