To the Moon, Mars, and Beyond: NASA’s Space Launch System
New technology drives the assembly of this state-of-the-art rocket BY Jeff Ding, Fredrick Michael, AND Jeffrey W. Sowards Reprinted with permission: The AWS Welding Journal Since 2011, the National Aeronautics and Space Administration (NASA) has been working on this nation’s next heavy launch space vehicle, the Space Launch System (SLS). The SLS is the most sophisticated and complex rocket system ever built in the history of this country’s long-running space program. This will propel astronauts first to the moon, then to Mars, and finally to deep space travel and beyond. The core stage is the backbone of the SLS rocket, towering 212 ft tall and 27.6 ft in diameter. This is comprised primarily of the liquid hydrogen tank, which is the rocket’s fuel, and the liquid oxygen tank, which is the oxidizer. The fuel and oxidizer will power the core stage main engines during vehicle ascent into orbit. Much of the core stage manufacturing and development work has taken place at the EM32 Advanced Welding Development Facility at the NASA George C. Marshall Space Flight Center, Huntsville, Ala., where some of the most sophisticated solid-state welding tools in the world are found. Here, many of the formidable welding tasks and challenges found — not only in core stage flight hardware, but also in other flight hardware welded and fabricated in the facility, including the Launch Vehicle Stage Adapter and the Orion Stage Adapter — have been addressed by a team of government engineers and welding technicians, as well as engineering support contractors, including the core stage prime contractor, Boeing. Fabrication and welding of the core stage hardware has been much more intense than the hardware that comprised the past External Tank Program, in that many of the SLS weld joints are thicker, more complex, and now entail the task of filling the exit holes left after the primary joining technique, self-reacting friction stir welding (SR-FSW). Therefore, SLS welding engineers have been tasked with solving unknown and unique welding challenges not seen before in rocket welding fabrication. A few of these challenges and current approaches to address them are presented in this story. Enabling Welding Technologies Solid-state welding processes are used almost exclusively in the fabrication of the 2000-series precipitation-hardened aluminum alloys used to manufacture SLS core stage cryogenic fuel tanks. The core stage hydrogen tank is shown in Fig. 1. The 2000-series aluminum alloys are used exclusively in the production of the cryogenic fuel tanks due to their excellent mechanical properties at the cryogenic temperatures of liquid hydrogen and liquid oxygen, as well as superior stress-corrosion cracking resistance, which is necessary for the humid saltwater environments present at NASA launch facilities. Solid-state welding techniques have been preferential since their introduction in the Space Shuttle program. They typically provide improved joint efficiencies (ratio of weld ultimate tensile strength to base metal ultimate tensile strength) when welding the 2000-series alloys as compared to fusion welding processes. Joint efficiency realized after optimized solid-state welding is typically ~ 75% as compared to only ~ 58% following conventional arc welding of, for example, AA2219 T87. Conventional friction stir welding, which has been used for many years in aerospace, is used for vertical welds in each of the five “barrel sections” comprising the liquid hydrogen tank (LHT). During LHT assembly, each barrel section is stacked together, and then joined with the SR-FSW process (Ref. 1). After the five-barrel sections are joined, forming a long cylinder, a ring and a dome is welded to each end of the cylindrical barrel section, thus forming the LHT. The liquid oxygen tank, although much smaller than the LHT, is welded in a similar fashion. The SR-FSW process, shown in Fig. 2, is a unique variation of FSW in that the reaction forces are self-contained by adding an independently controlled root shoulder that floats around the weld root face contour via force control on the root side of the weld. After completion of a joint produced via SR-FSW, a hole remains in the structure due to extraction of the pin tool. This hole is drilled out and filled by a plug welding process known as friction pull plug welding (FPPW) (Ref. 2), whereby a plug is inserted through the hole from the inner diameter of the core stage tank — see Fig. 2. Due to limited access inside the space vehicle structure and inability to place a backing anvil on the inner diameter, FPPW is required compared to the more conventional push plug welding, which requires a backing anvil. Current Efforts Improved Weld Schedule Transfer One of the major challenges when building large-scale space structures is ensuring the welding process parameters established in the development laboratory on weld panels are robust enough that similar joint properties are achieved when scaling to the large flight structure, such as the SLS. Weld process development is performed on panels that are a few feet in size with appropriate thickness to match the weld root face of the actual structure (most core stage tank weld root faces are less than 0.750 in. thick). However, there are also notable differences between the welding of development panels (approximately 24 in. long) and large flight articles in SLS. Panels are monocoque, whereas flight components contain internal structural stiffening features, such as isogrid or orthogrid. Also, the weld tooling and fixturing used for development is designed to closely match the flight welding tool, yet variations in stiffness between a clamped panel on a bench top and a full-scale structure are possible. These differences highlight two important aspects of the welding process that may be influenced. First, the effective heat conduction out of the weld can vary somewhat between a small panel and the large SLS structure. Second, the differences in self-restraint and external restraint on a small panel vs. a large barrel can result in different levels of weld shrinkage and distortion. After optimization of the SR-FSW and FPPW processes at the panel level, there is no intermediate step between a panel and full-scale article, i.e., the large-scale structures are typically welded with parameters developed at the panel level. It is impractical and cost prohibitive to weld large, flight-like aluminum structures during weld development. To give the welding operator flexibility, a “trim” range is usually allowed, whereby the welding operator can adjust weld parameters within a predefined window on-the-fly. While this approach has been successful for many years, new tools are available to approach the scaling problem with a new point of view as discussed below. Some work has been done to better characterize how a panel-level weld might differ from a flight structure in these regards by comparing in-situ measurements during panel welding and welding of a 28-ft-diameter barrel section. In-Situ Thermomechanical Metrology The accuracy, precision, and general usability of thermal imaging systems has greatly improved in recent years. Development of methods at NASA that couple 2D thermal measurements and multichannel strain instrumentation has resulted in an improvement in our understanding of the differences between heat flow and residual stress development in panel and barrel welds. Figure 3 shows some results of this developed capability. Infrared thermal imaging is compared between a panel weld and large-scale barrel weld. Multiple forward-looking infrared cameras and thermocouples are installed and placed at relevant positions for both test panel-level and large-scale welds. The high-definition data sets are captured in real time and postprocessed and analyzed for in-weld characterization, as well as weld-to-weld correlations and additional delta shifts. Physics-Based Modeling The high-definition thermal data acquired via thermography and thermocouple measurements, in addition to mechanical testing of properties, nondestructive examination, weld-tool feedback, and extensive high-frequency data sets, are all inputs into a maturing computational modeling program that seeks to fully characterize both fusion and solid-state FSW. Regarding delta-shifts analysis and modeling, Fig. 4 shows several facets of the modeling effort. Advanced computational efforts have been able to pinpoint and validate weld thermomechanical material flow properties’ sensitivities at the weld pin tools. In addition, these same computational efforts have confirmed stagnation flow zones leading to imperfect welds, as well as verifying and independently quantifying experimental sensitivities investigations. Furthermore, they have been able to provide large-scale and high-fidelity data windows and, therefore, provide a more complete insight into welding (whereas traditional testing has captured sparse data). The computational approaches promise to provide low-cost and fast methods for novel process development for differing aerospace alloys and processes, especially for NASA’s SLS/Artemis relevant alloys, configurations, and thicknesses across multiple machines. They can provide a root-causes forensics methodology, too. Just as importantly, the advanced computational modeling described above can help in the development of a long sought after welder’s handbook for solid-state welding. This handbook should be able to use initial welding input parameters and effectively predict as-welded mechanical properties from computing the thermal-mechanical history of the weld. In addition to the above-referenced solid-state FSW and fusion welding, computational and modeling tools currently available (or in rapid development by NASA and partners), there are now significant investments by NASA in additive manufacturing (AM) computational modeling, methods, and tools. Specifically, powder-bed selective laser melting, directed energy deposition, and solid-state AM are active additive manufacturing programs at NASA. Specifically, these AM technologies are focused on engines and structural components, rocket nozzles, in-space manufacturing, nuclear thermal propulsion (fuel element and reactor configurations), and fabrication of novel CERMET/refractory materials alloys. Therefore, multiple divisions and laboratories across several NASA centers are active in the development of AM technologies and the modeling of AM. Currently, state-of-the-art modeling includes the capability of modeling powder-bed alloy builds (prints) and the extraction of anisotropies due to beam scan travel directionality and anisotropies due to the microstructural properties of texturing, grains, and porosity. Another state-of-the art effort focuses on predicting mechanical residual stresses and deformations due to the deposition history; this is for “melt” printing. In the case of solid-state additive, such as in the mobile end-effector laser device deposition (wire, rod, or powder), collaborations with industry and research-leading partners are additionally providing investigational experimental data and new computational models for the solid-state additive methods. Cutting-edge, novel approaches that seek to take into account laser-scan path deposition are also in active collaborative investigation with NASA partners. Future challenges for the aerospace community are numerous. Several important hurdles will benefit from continued application of new science-based technologies, such as advanced metrology and computational modeling to better understand transfer of weld schedules from development to production. The transition from subtractive manufacturing to AM has made possible unique combinations of structure and chemistry; however, many additive builds need to be welded after printing. These 3D-printed parts are often treated as wrought material, yet experience in weld development on advanced propulsion alloys that have been 3D printed reveals they tend to behave much differently from wrought alloys from a weldability standpoint. Therefore, wrought materials should not be defaulted for weld development efforts without careful consideration. A second challenge is understanding how weld processes developed on Earth will behave in an in-space manufacturing environment (zero gravity and vacuum conditions), where heat and fluid transport caused by convection is often negligible. Extended missions to the lunar surface and beyond will likely necessitate in-space assembly and repair of metallic structures with fusion welding methods. It is certain that clever application of weld simulations will be necessary to enable critical in-space welds where weld experience and empirical data is limited. Conclusion The SLS is the most powerful booster the world has ever seen. One day soon, it will propel NASA astronauts and the agency’s Orion crew capsule on exciting missions — first to the moon, Mars, and then to deep space. For the first flight, hardware has been fabricated. Hardware has been stringently evaluated and tested. Ready to go. We are there. Fabrication of future hardware, however, may not go as planned. Due to the complexity of the hardware, some tests results may catch us by surprise. Untimely. Unexpected. Unknowns. One area where NASA and support engineers excel is dealing with the unexpected and unknowns, creating the necessary tools to understand, analyze, and overcome formidable challenges and achieve successful outcomes. Furthermore, this paper describes just a few of the advances in metrology and computational modeling relative to 3D printing, AM, and weld simulations and how these technologies can be exploited to advance NASA’s capabilities to increase the reliability and repeatability of not only SLS core stage initial welds but all welds on SLS and supporting hardware. Finally, the innovations described in the above text are intended to solve not only NASA obstacles and challenges, but are intended to be applied to other industry segments where the impact of NASA’s technology contributions can assist in numerous welding applications. References- Carter, R. W. Auto-adjustable tool for self-reacting and conventional friction stir welding. US 6758382 B1, United States Patent and Trademark Office, July 6, 2004.
- Coletta, E. R., and Cantrell, M. A. Friction pull plug welding: Chamfered heat sink pull plug design. US 6880743 B1, United States Patent and Trademark Office, July 19, 2005.
