2

Nuclear Thermal Propulsion

SYSTEM CONCEPT

A nuclear thermal propulsion (NTP) system is conceptually similar to a chemical propulsion system, where the combustion chamber has been replaced by a nuclear reactor to heat the propellant. Figure 2.1 depicts the basic components of an NTP system, which consists of three highly integrated subsystems: a nuclear reactor, a rocket engine, and a propellant storage and management subsystem. The reactor subsystem consists of the core, control drums and their actuators, reflector, shield, and pressure shell. The engine subsystem consists of the turbomachinery (including associated valves and pipes) and nozzle, and the liquid hydrogen (LH₂) tank and helium pressurization tanks are part of the propellant storage and management subsystem.

In both NTP and nuclear electrical propulsion systems (and terrestrial nuclear power plants), the reactor produces heat from fission of nuclear fuel.¹ Nuclear reactors also produce high levels of radiation that require shields to reduce the exposure of people and materials in the vicinity of the reactor. For an NTP system, the LH₂ propellant from the cryogenic LH₂ tank is delivered to the reactor using one or more turbopumps and the propellant management components. The LH₂ is directly heated by the nuclear reactor and then accelerates out the nozzle to generate thrust. This is in contrast to generating heat with combustion, as is the case in a chemical rocket. The control drums, which absorb neutrons, are situated around the outer annulus of the reactor core within the reflector. The drums are used to turn the reactor “on” and “off” and to increase or decrease reactor power. The hydrogen turbopumps are used to control the mass flow rate and pressure of the hydrogen propellant.

Figure 2.2 shows a reactor core cross section and fuel segment cluster of the NTP nuclear reactor of a type developed by the Rover and Nuclear Engine for Rocket Vehicle Applications (NERVA) programs.²Figure 2.2 depicts tightly packed hexagonal (also known as prismatic) fuel elements. This particular core is surrounded by 12 control drums, which are partially covered by reflector material and which reflect neutrons emitted from the core back into the core, to help sustain nuclear fission during reactor operation. Power is controlled in the reactor by rotating the drums. There is an inner and outer pressure vessel and reflector materials surrounding the control

___________________

¹ Although many isotopes of various elements can be used as nuclear fuel, uranium-235 (U-235) is the fuel of choice for all space nuclear propulsion designs under development by the United States.

² From 1955 until 1973 the Atomic Energy Commission’s Project Rover sought to develop nuclear reactors suitable for use with an NTP system. From 1961 until 1973 NASA’s NERVA Program sought to develop a complete NTP system. Both programs were jointly managed by the Space Nuclear Propulsion Office.

drums. Within each fuel element cluster are the tie tubes. The purpose of the tie tubes for the Rover/NERVA type cores is to regulate the temperature of the outer edge of the fuel elements and to provide some structural support to the fuel elements in the core.

HISTORICAL OVERVIEW

The Rover/NERVA reactor and NTP engine development program included a ground testing campaign that built and tested 22 reactors, using highly enriched uranium (HEU) graphite fuel with uranium dioxide (UO₂), uranium dicarbide (UC₂), and coated UC₂ particles.^3,4 In addition to demonstrating controlled reactor operation, the Rover/NERVA programs demonstrated the feasibility and challenges of the NTP engine concept. Specifically, the feasibility of using a nuclear reactor to heat the hydrogen propellant to generate predicted values for specific impulse (I_sp) using flow paths through solid graphite HEU fuel, tie tubes, and turbomachinery was demonstrated. The ground test campaign enabled the design of the Rover/NERVA reactors to iteratively evolve in response to issues identified during testing. For instance, the early Rover/NERVA reactors such as Kiwi B4A had structural issues caused by flow-induced vibrations that necessitated the testing and destruction of at least two entirely

___________________

³ J. L. Finseth, Rover Nuclear Rocket Engine Program: Overview of Rover Engine Tests. Final Report, Sverdrup Technology, Inc., Huntsville, AL, 1991, https://ntrs.nasa.gov/citations/19920005899.

⁴ HEU refers to uranium that is enriched to the point that 20 percent or more is U-235, which is fissile, with the remainder being uranium-238, which is not fissile. High-assay, low-enriched uranium (HALEU) refers to uranium that contains between 5 percent and 20 percent U-235. Naturally occurring uranium contains less than 1 percent U-235.

different reactors in order to isolate and fix the problem.⁵ Lessons were also learned regarding neutron moderation.⁶ Neutron moderation was primarily achieved by the graphite in the fuel elements, which operated simultaneously as a heat transfer element and as the primary neutron moderator.⁷

The Pewee reactor from the NERVA program incorporated additional, separately cooled zirconium hydride (ZrH) moderator material in its tie tubes and demonstrated a peak fuel temperature of 2750 K and a propellant temperature of 2550 K at the reactor exit, which corresponds to approximately 875 s I_sp in a vacuum with ideal expansion. Pewee accumulated two 20-min runs at full design power of 500 megawatt thermal (MWt), with a total of 192 min above 1 MWt. Most fuel elements used a niobium carbide (NbC) coating on graphite surfaces exposed to hot hydrogen, but a few fuel elements were coated with zirconium carbide (ZrC) instead. Pronounced cracking of the NbC graphite coating was observed, with significantly less deterioration for the ZrC coatings. The Pewee reactor also demonstrated that by adding the additional ZrH moderator material into an HEU core, the overall mass of the system could be decreased, making Pewee the smallest, highest-performing reactor in the NERVA series.⁸ However, the reactor life was unclear, and the fuel used in the Pewee reactor is not being considered for current NTP systems.⁹ The XE-Prime reactor from the NERVA program successfully demonstrated a record number of engine starts, shutdowns, and restarts that far exceeds the requirements of a Mars mission (i.e., 28 reactor starts, although some of the engine parts [e.g., turbopump bearings] had to be replaced).¹⁰

One of the keys to maximizing the I_sp of an operational NTP system is to shorten as much as possible the time it takes to start up and shut down the system. I_sp is directly related to operating temperature, so I_sp is reduced during reactor startup and shutdown when hydrogen propellant is flowing through the reactor but is not being heated to full operational temperature. In particular, the startup transient of the NTP reactor should allow the system to reach full operating temperature in 1 min or less in order to reduce the performance reduction for each individual engine firing, which would generally be less than 30 min each and can be as short as 10 min. These rapid transients introduce many design challenges throughout the system.

Table 2.1 summarizes the measured and predicted values (theoretical, assuming ideal conditions, and no losses) for I_sp for a sampling of reactors and engines tested after the preliminary Kiwi series.

The NTP performance requirements for the baseline mission require a maximum fuel temperature high enough to heat propellant to a temperature of approximately 2700 K at the reactor outlet (see Table 1.3).

Table 2.2 provides additional information on the maximum operating temperature of fuel forms used in historic NTP materials programs. The nonnuclear prototypic testing designation typically refers to furnace temperature testing that was prototypic or exceeded the fuel’s designed operating temperature. Nuclear testing typically denotes testing in a research reactor facility. Full-core testing signifies that the fuel was used as the primary or sole fuel source in a fully functioning nuclear core. As shown, several advanced fuel forms with greater than 2700 K performance have been produced and undergone environmental testing in high-temperature furnaces, in radiation fields, and in a combination of temperature and hydrogen exposure. Although the fuel forms listed in Table 2.2 have not been demonstrated under the integrated effects of an NTP engine operation, for the most part, they were able to withstand the maximum operating temperatures shown without exhibiting significant degradation.

While most of the Rover/NERVA research reactors did not use flight-configured engine hardware, there were a few reactors tested with NTP engine hardware components, with the XE-Prime being the system closest to

___________________

⁵ B.L. Pierce, Comparison of Analytical and Experimental Flow Induced Core Vibrations,” WANL-TME-645, Westinghouse Electric Corp., Pittsburgh, PA, 1964.

⁶ Neutron moderation is a broad term that refers to the effect a material has on lowering the energy of a neutron, such that the neutron’s energy is at an optimal level for capture by a fissionable material leading to nuclear fission. Sometimes a material is included in a reactor strictly for purposes of moderating neutrons, in effect making that material “the moderator” of the reactor. In the case of NERVA, the graphite in the fuel was not meant solely for the purpose of moderating neutrons (even though it had moderating effects); even so, graphite can be used by itself as a moderator in some reactor designs.

⁷ J.M. Taub, Review of Fuel Element Development for Nuclear Rocket Engines, LA 5931, Los Alamos Scientific Laboratory, Los Alamos, NM, 1975.

⁸ Moderators would likewise reduce the size of NTP reactors fueled by HALEU.

⁹ J.L. Finseth, Rover Nuclear Rocket Engine Program: Overview of Rover Engine Tests. Final Report, Sverdrup Technology, Inc., Huntsville, AL, 1991, https://ntrs.nasa.gov/citations/19920005899.

¹⁰ D.R. Koenig, Experience Gained from the Space Nuclear Rocket Program (Rover),” Los Alamos National Laboratory, 1986, https://fas.org/nuke/space/la-10062.pdf.

TABLE 2.1 A Sampling of Data from Reactor and Engine Tests That Occurred in the Later Stages of the Rover/Nuclear Engine for Rocket Vehicle Applications (NERVA) Programs

NOTE: In these tests, reactor fuels were exposed to the integrated effects of startup, operation, and shutdown through ground testing of a complete nuclear thermal propulsion engine configuration. The Pewee reactor demonstrated the hottest measured fuel and propellant exit temperature of the Rover/NERVA series. The NRX Engine System Test and XE-Prime reactors were both tested engine hardware which was the closest to being “flight-like.” All of these tests were conducted between 1964 and 1969, inclusive.

SOURCE: Adapted from J.L. Finseth, Rover Nuclear Rocket Engine Program: Overview of Rover Engine Tests. Final Report, Sverdrup Technology, Inc., Huntsville, AL, 1991, https://ntrs.nasa.gov/citations/19920005899; and D.R. Koenig, Experience Gained from the Space Nuclear Rocket Program (Rover),” Los Alamos National Laboratory, 1986, https://fas.org/nuke/space/la-10062.pdf.

the envisioned operational system.¹¹ This experimental engine test incorporated pump and hardware arranged as designed for flight (i.e., close-coupled propellant feed system similar to the reactor and engine hardware arrangement seen in Figure 2.1), although it was only tested to 710 s I_sp. Additionally, the NRX/EST was an engine system test that used a breadboard that connected relevant flight hardware to the reactor while mounted to a train car.

Although the Rover/NERVA programs demonstrated proof of concept for an NTP system, the programs were cancelled before program goals were achieved due to a shift in funding priorities. Consequently, no complete NTP system has been assembled and tested in its flight configuration or flown in space. Other NTP programs have been carried out since Rover/NERVA, but none have built any additional reactors or engines. The Argonne National Laboratory (ANL) and General Electric GE-710 programs developed concepts for fast-spectrum¹² ceramic-metal (cermet) fuels for nuclear-powered aircraft and NTP concepts that utilized HEU.¹³ Cermet fuels, such as tungsten uranium dioxide (WUO₂), were manufactured and tested. The Space Nuclear Thermal Propulsion (SNTP) program was primarily a fuel development effort for the particle bed reactor that tested the use of coated HEU particles for NTP, and it identified many challenges. The SNTP program also conducted moderator block experiments using polyethylene moderator material,¹⁴ and it produced hardware for nonnuclear component engine testing. Ground testing of complete SNTP reactors was planned, but not implemented, before program termination. The Soviet Union had NTP development efforts as well (such as the RD-410), which purportedly used a unique (twisted ribbon) carbide HEU fuel and a ZrH moderator.^15,16

___________________

¹¹ D. Sikorski and R.T. Wood, “Nuclear Thermal Rocket Control,” Proceedings of American Nuclear Society Topical Meeting: Nuclear and Emerging Technologies for Space, 2019, http://anstd.ans.org.

¹² A fast-spectrum reactor is designed to rely predominately on fast (unmoderated) neutrons, whereas a thermal-spectrum reactor relies predominantly on moderated (thermal neutrons). Fast-spectrum reactors, which require a more intense radiation field, can be designed to use HALEU, but they are more compatible with HEU because it has a higher concentration of fissionable uranium (i.e., U-235) relative to HALEU. Thermal-spectrum reactors can be designed to use either HEU or HALEU.

¹³ S.K. Bhattacharyya, An Assessment of Fuels for Nuclear Thermal Propulsion, ANL/TD/TM01-22, Argonne National Laboratory, Lemont, IL, 2002, https://www.osti.gov/servlets/purl/822135.

¹⁴ R.A. Haslett, Space Nuclear Thermal Propulsion Program, Grumman Aerospace Corp., Bethpage, NY, 1995.

¹⁵ Z. Vadim and V. Pavshook, “Russian Nuclear Rocket Engine Design for Mars Exploration,” in Tsinghua Science and Technology 12(3): 256-260, 2007.

¹⁶ A. Lanin, “Nuclear Rocket Engine Reactor,” Springer Series in Materials Science, Vol. 170, Springer-Verlag Berlin Heidelberg, Germany, 2012, doi:10.1007/978-3-642-32430-7.

STATE OF THE ART

This section discusses the state of the art of the subsystem technologies that make up an NTP system as well as associated modeling and simulation (M&S) capabilities.

Reactor Subsystem

The only data available in the United States that can be used to validate NTP reactor models are from HEU reactor-engine subsystems in the 1960s and 1970s; there are no experimental data on high-assay, low-enriched uranium (HALEU) NTP subsystems.

The current state of the art for the reactor subsystem is limited to the M&S capabilities used to analyze a reactor virtually. Existing hardware manufacturing capabilities are insufficient to build an NTP reactor at the scale required for cargo or crewed missions associated with the baseline mission. Current M&S capabilities can generate steady-state neutronic designs of NTP reactors to simulate the nuclear core sustaining a chain reaction.^17,18 Reactor core models can be coupled with thermal-hydraulic, fluid models for simplified one-dimensional core-wide approximations and higher fidelity simulations for subscale analyses (i.e., using computational fluid dynamics simulations).^19,20 Numerous M&S design studies derive new concepts based on prior NERVA-type experiments. However, the coupled neutronic, thermal-hydraulic and engine balance of plant M&S tools are limited in their ability to reliably model NTP systems for which there are no experimental data for model validation, particularly for transients. Additionally, state-of-the-art M&S tools lack the ability to conduct coupled, high-fidelity analyses to assess system lifetime and potential failure mechanisms. Modeling the dynamic nature of the nuclear reactor, coupled with the flow of propellant and change of temperature, has not been completed for NTP, and significant uncertainty remains in the materials interactions between the hydrogen propellant and the reactor fuel. Simulations of dynamic reactor behavior exist, such as the dynamic modeling capability of Los Alamos National Laboratory that was used for the Kilopower program’s subscale power system test.²¹ Such tools will need to be adapted and benchmarked against test data, for NTP dynamic modeling, which use different materials under different conditions, different scales, and different working fluids. State-of-the-art M&S tools lack the capability for mechanical and structural simulation of reactors needed to assess the potential for flow-induced vibration issues, such as those faced by Kiwi B4A. This is also due to a lack of materials data that are needed for M&S inputs, such as for block monolithic ZrH at elevated temperatures.

Fuels

Several ceramic composite fuel forms (ceramic fuel particles in a graphite matrix with a protective NbC or ZrC coating) were demonstrated in the NERVA program to exhibit acceptable behavior in flowing hydrogen, mostly up to propellant temperatures of about 2550 K during reactor testing, with the Pewee fuel setting a record at 2750 K at its peak. Cermet fuels were not reactor-tested in the NERVA program, but thermal cycling tests demonstrated a mass loss of less than 1 percent for WUO₂ cermets up to 2800 to 3000 K in flowing hydrogen for 70 to 193 thermal cycles.²² NASA, the Department of Energy (DOE), and other research organizations, including those in industry and academia, have attempted to build solid HEU-based fuel rods. These efforts have included fuel rods comprised of graphite (i.e., Rover/NERVA derived), cermets (i.e., ANL/GE-710 derived), and other refractory blends.^{23,24,25,26,27} The fuel types are largely derivative of a variety of historic HEU reactor core concepts. These

___________________

¹⁷ Los Alamos National Laboratory, “Monte-Carlo N Particle (MCNP) Transport Code,” https://mcnp.lanl.gov/, accessed May 22, 2021.

¹⁸ Idaho National Laboratory, “Multiphysics Object-Oriented Simulation Environment (MOOSE), An Open-source, Parallel Finite Element Framework,” https://www.mooseframework.org/, accessed May 22, 2021.

¹⁹ Siemens, “Simcenter STAR CCM+, Engineer Innovation with Multiphysics Computational Fluid Dynamics (CFD) Simulation,” https://www.plm.automation.siemens.com/global/en/products/simcenter/STAR-CCM.html, accessed May 22, 2021.

²⁰ Ansys, Inc, “Ansys Fluent, Fluid Simulation Software,” https://www.ansys.com/products/fluids/ansys-fluent, accessed May 22, 2021.

²¹ D.I. Poston, M.A. Gibson, R.G. Sanchez, and P.R. McClure, “Results of the KRUSTY Nuclear System Test,” Nuclear Technology 206(sup1): 89-117, 2020, doi: 10.1080/00295450.2020.1730673.

²² M.E.M. Stewart, and B.G. Schnitzler, “A Comparison of Materials Issues for Cermet and Graphite-Based NTP Fuels,” Proceedings of 49th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 2013, https://arc.aiaa.org/doi/book/10.2514/MJPC13.

²³ R.C. O’Brien, N.D. Jerred, S.D. Howe, R. Samborsky, D. Brasuell, and A. Zillmer, “Recent Research Activities at the Center for Space Nuclear Research in Support of the Development of Nuclear Thermal Rocket Propulsion,” 2012 Nuclear and Emerging Technologies for Space (NETS 2012). INL/CON-10-19825, 2012, https://www.lpi.usra.edu/meetings/nets2012/pdf/3060.pdf.

²⁴ M.W. Barnes, et al., “NTP CERMET Fuel Development Status,” 2017 Nuclear and Emerging Technologies for Space (NETS 2017), American Nuclear Society, La Grange Park, IL, 2017.

²⁵ M.W. Barnes, D.S. Tucker, and K.M. Benensky, “Demonstration of Subscale Cermet Fuel Specimen Fabrication Approach Using Spark Plasma Sintering and Diffusion Bonding,” in 2018 Nuclear and Emerging Technologies for Space (NETS 2018), American Nuclear Society, La Grange Park, IL, 2018.

²⁶ B. Jolly, M. Trammell, and A.L. Qualls, “Coating Development on Graphite-Based Composite Fuel for Nuclear Thermal Propulsion” in Proceedings of the 51st AIAA/SAE/ASEE Joint Propulsion Conference, 2015, doi:10.2514/6.2015-3777.

²⁷ S. Raj, J. Nesbitt, and M. Stewart, “Development of Advanced Coatings for NERVA-Type Fuel Elements,” in 2015 Nuclear and Emerging Technologies for Space (NETS 2015), http://anstd.ans.org/wp-content/uploads/2015/07/3006.pdf.

solid fuel elements have been created using hot isostatic pressing, spark plasma sintering, and other methods, and have undergone testing in facilities such as the Compact Fuel Element Environment Simulator and the Nuclear Thermal Rocket Element Environment Simulator, which can achieve isothermal, steady-state temperatures in excess of 2500 K in the presence of flowing hydrogen.

A driving characteristic for the design of NTP systems is the high operating temperatures in the reactor core; an I_sp of 900 s corresponds to a hydrogen propellant reactor exit temperature of approximately 2700 K.^28,29

NASA is currently involved in testing uranium nitride (UN) cermet and ceramic-ceramic (cercer) fuel forms, including high-temperature hydrogen testing of uncoated fuels, and is planning further nonnuclear and nuclear testing. Upcoming nonnuclear prototypic testing will include flowing hot-hydrogen furnace testing at temperatures greater than or equal to 2850 K of the following: tungsten-coated UN particles, ZrC-coated UN particles, tungsten/molybdenum alloy-UN cermet composite fuel, and ZrC-UN cercer composite fuel, as well as full length cermet and cercer fuel elements.

NASA has also indicated an interest in solid-solution carbide fuel technology with coated carbide particles.³⁰ Although the United States has not successfully demonstrated NTP solid-solution quaternary carbide fuel forms, there is some limited documentation on the Russian RD-410 NTP fuel technology.³¹ Because the melting temperatures of UC₂ and uranium carbide (UC) particles are 2730 K and 2780 K, respectively,³² it is difficult for an NTP system to achieve a propellant reactor exit temperature of approximately 2700 K with uncoated carbide particles, unless they are mixed in a solid solution with refractory carbides such as ZrC.

Moderators and Non-Fuel Materials

Limited HEU reactor subsystem data exist on the successful performance of ZrH moderators in the single-pass tie tubes of the Pewee HEU reactor. Issues such as power oscillations,³³ hydrogen migration and hydrogen dissociation, and loss at temperatures above 700 K remain as potential challenges for the incorporation of ZrH into an NTP reactor. ZrH moderator blocks were demonstrated in Soviet Thermionic Operating Reactor Active Zone (TOPAZ) space reactors. Two TOPAZ I reactors were launched as flight demonstrations, and in the early 1990s the United States purchased a developmental TOPAZ II reactor for nonnuclear test and evaluation.^34,35,36 In addition, DOE is currently manufacturing and testing ZrH. The United States, however, has no flight experience with moderator block technology.

Beryllium (Be) is often proposed for use in NTP designs as a reflector and as a moderator. Beryllium can be used in forms such as beryllium oxide (BeO) or in pure form (Be). It is most suitable for components with an operating temperature of less than 1000 K; reactors and cooling approaches are designed to ensure this temperature is not exceeded. Beryllium was used for the reflector and in the control drums for the NERVA reactors and in a general capacity for a variety of other nuclear power reactors.³⁷

___________________

²⁸ C.R. Joyner, et al., “NTP & NEP Design Attributes for Mars Missions,” presentation to the committee, virtual meeting, June 29, 2020, p. 8.

²⁹ D. Burns, “DOE Role in Nuclear Thermal Propulsion Technology Development,” presentation to the committee, virtual meeting, June 22, 2020.

³⁰ M. Houts, “Nuclear Thermal Propulsions,” presentation to the committee, virtual meeting, June 8, 2020.

³¹ Z. Vadim and V. Pavshook, “Russian Nuclear Rocket Engine Design for Mars Exploration,” in Tsinghua Science and Technology 12(3): 256-260.

³² D. Manara, et al., “High Temperature Radiance Spectroscopy Measurements of Solid and Liquid Uranium and Plutonium Carbides,” Journal of Nuclear Matter 426: 126-138.

³³ D.S. Stafford, “Multidimensional Simulations of Hydrides During Fuel Rod Lifecycle,” Journal of Nuclear Materials 466: 362-372, 2015.

³⁴ D. Buden, Summary of Space Nuclear Reactor Power Systems (1983-1992), Idaho National Engineering Laboratory, Idaho Falls, ID, 1993, https://www.osti.gov/servlets/purl/10151265.

³⁵ M.S. El-Genk, “Deployment History and Design Considerations for Space Reactor Power Systems,” Acta Astronautica 64(9-10): 833-849, 2009.

³⁶ V.N. Adrianov, et al., Topaz-2 NPP Reactor Unit Mechanical Tests Summary Report Vol. 1, CDBMB through INERKTEK Technical Report, Moscow, Russia.

³⁷ J.L. Finseth, Rover Nuclear Rocket Engine Program: Overview of Rover Engine Tests. Final Report, Sverdrup Technology, Inc., Huntsville, AL, 1991, https://ntrs.nasa.gov/citations/19920005899.

Engine Subsystem

Engine hardware, such as turbomachinery and valves, has evolved independent of NTP reactor hardware for use on chemical propulsion systems. Existing chemical propulsion engine components can be scaled, modeled, and integrated for NTP use. For instance, the RL-10 and similar turbopumps have been modeled for decades for a variety of NTP design studies while having undergone maturation and hardware testing for a variety of chemical propulsion uses, including in space.^38,39

M&S capabilities applicable to the nonnuclear (i.e., chemical) engine subsystem elements are well developed for both static and dynamic engine flow conditions. Some of these models^40,41 are also applicable to NTP engine subsystems and have been used to model both HEU and HALEU-type engines.

Propellant Storage and Management Subsystem

Long-term storage and active cryogenic technologies for LH₂ have similarly evolved independently of NTP, but significant challenges must still be overcome to meet a storage time of perhaps 4 years for the baseline mission (2 years in an assembly plus 2 years for the round trip to Mars). Ongoing research technology development by NASA will lead to several missions beginning in 2021 to demonstrate advanced technologies for the storage and transfer of cryogenic fluids in space.

TECHNOLOGY REQUIREMENTS, RISKS, AND OPTIONS

NTP system performance is strongly driven by the heat transfer efficiency of a given design. This is a function of the temperature profile during operations, time at the maximum operating temperature, the number of planned operating cycles (with safety margins for additional potential cycles), and rates of change for temperatures across the system. The primary NTP system-level risks are driven by the following:

• The high operating power density and temperature of the reactor necessary to heat the propellant to approximately 2700 K at the reactor exit for the duration of each burn. This is necessary to meet the 900 s I_sp mission requirement.
• The need for long-term storage and management of cryogenic LH₂ propellant.
• The much shorter NTP reactor startup times (as little as 60 s from zero to full power) relative to other space or terrestrial power reactors (sometimes as long as several hours).
• The longer startup and shutdown transients of an NTP system relative to chemical engines. This drives design of the engine turbopumps and thermal management of the reactor subsystem.

Reactor Subsystem

An NTP system with a propellant reactor exit temperature of approximately 2700 K represents an extreme environment in terms of temperature and hydrogen corrosion for the materials in the reactor core. This reactor operating temperature implies that there are few viable fuel architectures. The fuel element, which includes the fuel and cladding, the fuel assemblies, moderator, support structures, and the reactor pressure vessel must maintain physical integrity while cycled through the thermomechanical stress induced during repeated cycles of reactor startup, operation at power, shutdown, and restart.

___________________

³⁸ C.R. Joyner, et al., “LEU NTP Engine System Trades and Mission Options,” 2019 Nuclear and Emerging Technologies for Space, (NETS 2019), 2019.

³⁹ C.R. Joyner, et al., “NTP & NEP Design Attributes for Mars Missions,” presentation to the Space committee, virtual meeting, June 29, 2020, p.8.

⁴⁰ NASA, “NPSS Numerical Propulsion System Simulation,” https://software.nasa.gov/software/LEW-17051-1, accessed May 22, 2021.

⁴¹ NASA, “ROCet Engine Transient Simulation Software (ROCETS),” https://software.nasa.gov/software/MFS-31858-1, accessed May 22, 2021.

At least three new NTP fuel architectures are under consideration by NASA, including the following:

• Cercer-coated fuel particles in a refractory ceramic matrix,
• Cercer solid solutions of mixed (U, Zr, Nb)C fuel with multiple potential particles (UN, UC, UCZr, etc.), and
• Cermet-coated fuel particles in a refractory metal matrix.

Multiple fuel particle packing densities (15 to 70 volume percent) and varying fuel particle architectures are under consideration for some of these fuel options. An overview of these NTP fuel options is provided in Figure 2.3. Both HEU and HALEU fuel enrichments are possible. Currently, the reference cermet fuel architecture uses UN particle fuel at 40 to 70 volume percent packing density, with fuel particle architecture to be finalized, and a molybdenum (Mo)-30%W⁴² metal matrix. The metal matrix composition involves a compromise between limiting parasitic thermal neutron absorption (i.e., by reducing tungsten content) and maximizing the alloy melting temperature (i.e., by increasing tungsten content). Cercer fuels with coated fuel particles offer the potential for increased safety margins with respect to fuel matrix melting compared to cermet systems, but cercer fuels are at a lower level of technological and fabrication maturity. Cercer solid solution fuels similarly offer the potential for higher performance and safety (fuel melting) margins but are at a similar lower level of technological maturity. Graphite matrix fuel systems have demonstrated excellent high-temperature capability (greater than 3000 K) but would require a robust, defect-free high-temperature coating such as ZrC for all surfaces exposed to hot hydrogen due to the fundamental high-temperature incompatibility of graphite and hydrogen. Pronounced cracking was observed in ZrC coatings on graphite composite fuel coolant channel surfaces even at temperatures as low as about 1500 K in the NERVA program, although more recent research has made advances in this area.⁴³

NASA and DOE will need to determine if current or planned HEU or HALEU fuel feedstock production capabilities will be sufficient to meet the needs of the NTP baseline mission. Key issues include identification of a suitable fuel architecture. Trade studies to address these issues will be needed in advance of mission formulation and initial design efforts.

Testing of candidate core materials may consider the applications of core fabrication for both conventional and advanced manufacturing methods. Advanced methods, such as additive manufacturing, are showing promise in both aerospace and nuclear manufacturing industries. These techniques are most likely to be suitable for NTP components outside the highest temperature environments of the reactor core. New manufacturing techniques, however, lack a substantive body of relevant performance testing in either nuclear or in-space high-radiation environments. As a result, new manufacturing techniques will require performance testing and analysis, even if these techniques are used to fabricate nuclear-qualified materials previously made using conventional techniques.

While the NTP propulsion concept has been studied for more than six decades, insufficient technical maturity for fuel forms, fuel assemblies, moderator materials, and high-temperature structural materials, in particular for the three new core concepts noted above (cermet, cercer, and cercer/carbide), is a significant risk to overall program success. Without substantial up-front investment in the development of these specific areas of technical risk, many of the integrated system designs and associated integrated risks cannot be adequately managed or mitigated.

Thermodynamically stable high-performance neutron moderators are an important aspect of a thermal-spectrum NTP reactor core design. Potential moderator materials include ZrH, YH, Be, BeO, and Be₂C. For the hydrides and pure beryllium, maximum use temperatures are expected to be about 700-1500 K due to hydrogen dissociation and Be melting concerns. Upper operating temperature limits based on dissociation need to be accurately determined so that candidate moderator materials can be assessed and cooling channels designed. The effects of hydrogen embrittlement and infiltration into the moderator material and the related dissociation characteristics are also important considerations. These characteristics will need to accommodate temperature and thermomechanical excursions experienced during startup and shutdown transients of the NTP reactor system. In particular, hydrogen

___________________

⁴² That is, an alloy of molybdenum with 30 percent by weight of tungsten.

⁴³ S.V. Raj and J.A. Nesbitt, “Development of Advanced Coatings for NERVA-type Fuel Elements,” 2015 Nuclear and Emerging Technologies for Space (NETS 2015), Albuquerque, NM, February 2015.

flow will need to continue after reactor shutdown to provide cooling, which will impose some performance penalty on the NTP system.

The reactor structure serves as the primary interface to the LH₂ propellant at the upper plenum inlet as well as the interface to the nozzle. While there has been some progress in the development of reactor structural materials and the initial design of these systems, key aspects of the reactor structure will need to be matured to accommodate optimization for mass, propellant flow, propellant pressure drop through the length of the structure, flow rates and stability, and stress analysis. Shielding, for both gamma and neutron radiation, will also need to be considered as part of the overall reactor system design. Shielding design, composition, and placement will need to account for the location of instrumentation and control electronics, radiation susceptible turbomachinery, the cryotank for LH₂ storage, and the crew vessel.

Reactivity control (i.e., thermal neutron absorber) materials are also needed for operational control, launch safety considerations, and multiple reactor startups and shutdowns. Many of the existing and current designs employ reactivity control drums located radially within a reflector assembly. Control drums are primarily constructed of a neutron reflector material with a section of a thermal neutron absorber, such as B₄C, that can be rotated to face the core (to shut down the reactor) or rotated away from the core (for reactor startup and operation). The control mechanism for the control drums will need to manage the flow rates for LH₂ and gaseous hydrogen (H₂) through the moderator, outer reactor structure, nozzle, and reactor core. For the initial flight of an NTP system, additional sensors for temperature, pressure, coolant flow rate, and neutron flux will likely be necessary to provide additional characterization of the flight system.

Engine Subsystem

The engine subsystem has significant heritage from chemical rocket engines, including the use of gaseous H₂ and LH₂ as a fuel. Additional testing for the engine subsystem will be necessary to demonstrate integrated operability, lifetime, and reliability. However, assuring the performance of the engine subsystem is a relatively low-risk element of developing an NTP system for the baseline mission.

Propellant Storage and Management Subsystem

The development of multiyear cryogenic storage capabilities for LH₂ remains a significant challenge. Storage of metric tons of LH₂ at cryogenic temperatures as low as 20 K, with minimal losses, is needed because of the long duration of the baseline mission, including time for in-space vehicle assembly and the round trip to Mars. The current expectation for the baseline mission is that at least six NTP system starts will be needed, with a total LH₂ propellant requirement that ranges from 7 to 21 10,000-kg tanks of LH₂ depending on which launch vehicles are used and the mission departure year. Minimizing the boiloff of LH₂ from the storage tanks is necessary to reduce the amount of LH₂ that must be launched and the number of storage tanks that must be integrated into the Mars exploration spacecraft.⁴⁴

Although development of refrigeration technology is proceeding, existing cryocooling systems cannot reliably meet propellant tank requirements over a mission of this duration. Additionally, propellant mass must be accurately measured before and after each firing of the propulsion system to appropriately balance flow rate to the reactor startup and reactivity control operations. Cryocooling systems will require electrical power throughout the mission, which would be provided by small solar arrays that are dedicated to this purpose.

TESTING, MODELING, AND SIMULATION

As described above, the components of the engine subsystem and the propellant storage and management subsystem have been demonstrated on chemical rockets to a high technology readiness level (with the exception of long-term storage of LH₂ in space with minimal boiloff). Therefore, the largest return on testing, and on M&S efforts, would accrue through a focus on the reactor subsystem, wherein lies the dominant system risks.

Testing is conducted to verify material characteristics, operational performance, and functionality (e.g., operability, controllability, and thermal management) of components, subsystems, and integrated systems over the intended operational lifetime of the system, including transients and margins of safety. All previous rocket engines have undergone extensive, multi-engine full-scale ground testing as part of their certification programs. For example, 10 or more space shuttle main engines as well as J-2 and RL-10 upper-stage engines were all ground

___________________

⁴⁴ C.R. Joyner, et al., “NTP & NEP Design Attributes for Mars Missions,” presentation to the committee, virtual meeting, June 29, 2020, p. 8.

tested for more than their full mission durations prior to certification.⁴⁵ These tests also support retirement of potential concerns related to safety and reliability of both nuclear and nonnuclear elements of the NTP system, and acquired test data provide a means to validate models used to support computational design and simulation of system operation during both steady-state and transient conditions to ensure a sufficient level of confidence regarding design margins and uncertainty under all operational conditions.

A traditional progression of tests includes separate effects testing to characterize materials properties and behaviors; component, subassembly, and assembly testing; scaled system testing; and integrated system tests. For the reactor subsystem, this could entail testing of components in environment chambers or test reactors, subcritical tests, and critical tests (at zero-power and full-power conditions).

An NTP reactor core and associated systems present unique challenges relative to terrestrial nuclear technology employed for power production. NTP systems operate at much higher power density levels, temperatures, and coolant (propellant) flow rates than standard reactor technologies. They therefore use materials for which there is a very limited database that can be used for model validation. Additionally, in contrast to terrestrial power reactors, NTP reactors are open cycle: the reactor coolant is expelled through the nozzle rather than being contained in a closed cycle. Prior to establishing a test plan, the existing database of materials properties, under both steady-state and dynamic conditions, would be assessed, and a review of data available from previous test programs on related technologies (e.g., Rover/NERVA) would be conducted, including their compatibility with flowing H₂ at operational temperatures. If similar materials and operating parameters are selected, using available data to benchmark modern M&S tools may narrow the remaining areas of uncertainty, allowing developers to reduce the overall number and types of tests necessary to retire risk. In addition, testing of reactivity control for NTP systems has only been conducted for HEU systems, and no full engine tests have been completed. Testing is needed to characterize reactivity control of a moderated HALEU-fueled NTP system.

For an NTP reactor, the reactor is ramped to full power over a period of approximately 60 s while hydrogen propellant is introduced to the outer reactor containment vessel, and subsequently the core, for temperature control (i.e., cooling using LH₂). During the initial few seconds, thermal-mechanical stresses expand the reactor core, having a reactivity impact. Introduction of H₂ also has a reactivity impact. That is, multiple feedback effects occur concurrently and locally, such that the power increase may be nonlinear and scale dependent, making it difficult to predict and control unless these behaviors are well understood and represented in the corresponding dynamic simulation of the reactor startup. These interplaying conditions must be managed at high temporal fidelity to offset transient excursions in the reactivity profile. Passage of the H₂ propellant through the core may also introduce substantial core pressure variations, both axial and radial temperature variations, flow instabilities, and engine vibration, all of which are scale dependent. Finally, the pressure differential at the nozzle throat may also induce loads or pressure gradients that impact engine performance and safety. This last effect may be less important to characterize via ground test and may instead be characterized during initial cargo missions. Beyond these nuclear and thermal-hydraulic challenges, numerous thermomechanical verification challenges exist for these engines that will operate near the limits of temperature and material property and joining technology capabilities. It is critical to recognize that most of the complex interactions described above are nonlinear and scale-dependent, meaning that the risks they represent cannot be retired by subscale testing.

Ground tests of integrated NTP reactor and engine subsystems would reduce technical risk. Such testing has been used for all previous liquid rocket engines for flight.⁴⁶ While it may be possible to characterize integrated system performance using a nonnuclear, electrically heated environment, the accuracy of such testing may be a challenge for NTP systems, which have tightly coupled neutronic-thermal-hydraulic response characteristics. In addition, heating elements used to emulate nuclear heat in these tests would need to be designed to accurately

___________________

⁴⁵ S. Richards, “Liquid Rocket Engine Flight Certification,” Proceedings of Space Transportation Technology Symposium, Pennsylvania State University, 1991, https://ntrs.nasa.gov/api/citations/19910018936/downloads/19910018936.pdf.

⁴⁶ S. Richards, “Liquid Rocket Engine Flight Certification,” Proceedings of Space Transportation Technology Symposium, Pennsylvania State University, 1991, https://ntrs.nasa.gov/api/citations/19910018936/downloads/19910018936.pdf.

reflect core temperature profiles (both radially and axially) and heat-up/cool-down rates to adequately reflect both steady-state and dynamic operation.^47,48

A series of separate effects testing for materials, components, and subassemblies is necessary prior to selection of options and system-level testing. Specific tests to characterize properties and behaviors, both pre- and post-irradiation, would be determined based on materials selections and the existing databases associated with the selected materials. Scaled testing of subsystems, designed at a scale that provides performance results that support characterization of integrated effects (employing nondimensional parameters) and identification of potential failure mechanisms, may also be performed to further develop an understanding of system-level performance parameters.

Following the separate effects, component, and subassembly ground testing, system-level nuclear ground testing phases would take place. These tests could include those described below, which would be performed sequentially. Key considerations for testing in all three phases include potential requirements for safety and environmental reviews and approvals, especially for testing that requires the construction of new facilities or modification of existing facilities.

• Phase 1. Zero-power critical (ZPC) and low-power tests. ZPC tests are neutronic tests that verify various operational characteristics of a fission reactor. The ZPC test series is conducted in such a way that it leaves the reactor and components essentially nonradioactive. This method of testing would verify the operability of the reactivity control system but would not ramp the reactor through the full transient conditions to achieve full power. Hence, the ZPC approach would not demonstrate the thermomechanical stability of the reactor system, nor would it demonstrate the effects of the LH₂ propellant on reactivity control, system performance, and safety. All reactor designs would be subjected to ZPC testing, and possibly low-power testing, prior to developing a flight system, regardless of the decision to include other integrated system ground testing. A ZPC test would also be conducted for the flight unit prior to launch of an NTP for either a cargo or crewed mission.
• Phase 2. Reactor operational tests (Rover/NERVA-like testing). Operational testing of a complete nuclear reactor subsystem would entail nuclear testing of the complete, prototypic reactor system with heat generated by fission. LH₂ would be pumped through the reactor structure and core during startup, operation, and shutdown, as a demonstration of the engine system and thermal management system for the reactor through all phases of operation. Such an approach would demonstrate reactor operability, performance, reliability and, most importantly, controllability through transient startup, operation, and shutdown conditions, and it would demonstrate performance after multiple restarts. Properly instrumented, these operational tests would provide the validation data necessary to benchmark and demonstrate the efficacy of M&S tools in predicting reactor performance, lifetime, and reliability and characterizing hydrogen effects on the reactor materials, thermal management, and reactivity controls. These tests would also allow detailed post-test inspections to determine material effects and degradation and identify incipient failure mechanisms to allow for reactor-to-reactor manufacturing variability. While such tests would not incorporate the engine subsystem, they would require a test support system to manage the hydrogen effluent as it exits the reactor core. Facilities at the Nevada Test Site could be evaluated for their ability to support operational tests, but the need to capture and/or recirculate the H₂ coolant/propellant may make it difficult to use existing facilities without extensive modifications. Safety and environmental approvals will also be required.⁴⁹
• Phase 3. Integrated system tests. These tests would add the engine and propellant management subsystems to the reactor test configuration described in Phase 2.⁵⁰ The integrated system tests could be completed on

___________________

⁴⁷ S.M. Bragg-Sitton, R. Dickens, D. Dixon, R. Reid, M. Adams, and J. Davis, “Advanced Thermal Simulator Testing: Thermal Analysis and Test Results,” Proceedings of Space Technology and Applications International Forum (STAIF 2008), Albuquerque, NM, 2008, doi: 10.1063/1.2844995.

⁴⁸ S.M. Bragg-Sitton, R. Dickens, D. Dixon, R. Kapernick, M. Adams, and Joe Davis, “Development of High Fidelity, Fuel-Like Thermal Simulators for Non-Nuclear Testing,” Proceedings of Space Technology and Applications International Forum (STAIF 2007), Albuquerque, NM, 2007, doi: 10.1063/1.2437499.

⁴⁹ S.K. Borowski, R.J. Sefcik, J.E. Fittje, D.R. McCurdy, A.L. Qualls, B.G. Schnitzler, J. Werner, A. Weitzberg, and C.R. Joyner, Affordable Development and Demonstration of a Small Nuclear Thermal Rocket (NTR) and Stage: How Small is Big Enough?, NASA/TM- 2016-219402, AIAA-2015-4524, 2016, https://doi.org/10.2514/6.2015-4524.

⁵⁰ The integrated system tests would validate the complete NTP system with the exception of the ability for long-term storage of LH₂; those technologies can be tested separately.

• the ground, but they would require extensive investments in infrastructure and environmental approvals. Facilities to support these tests do not currently exist.

The ZPC tests described in Phase 1 can likely be conducted at existing facilities, including the launch site, although facility modifications will be needed.⁵¹ These tests can be used as a part of the design process and, when conducted on the flight unit, can verify neutronic status and control drum operability prior to system launch. Phase 1 testing would have the most modest schedule impact and cost relative to the other test phases, but ZPC testing would not retire many of the risks associated with dynamic performance of the system, particularly during startup to full power or during the shutdown transient.

The ground testing approach described in Phase 2 emulates that which was adopted for Rover/NERVA. Facilities used to support those historical tests are no longer available, but existing facilities could be modified to support the testing needs, given sufficient time and funding. Testing with this approach is necessary to fully understand the dynamic system performance, lifetime limitations, reliability, interfaces, and manufacturing margins, thus reducing uncertainty and risk to program success. To have the greatest impact on risk reduction, multiple test units would be needed to determine the repeatability of the measured system characteristics to properly assess design margins. This test series would only be initiated after (1) a thorough review of historical data to benchmark M&S codes against prior tests to characterize the most significant areas of uncertainty and potential failure modes and (2) a detailed subsystem testing campaign.

Legacy fuels, materials, and structural design approaches (e.g., from the Rover/NERVA program) could be used to mitigate some schedule and technical risk associated with an NTP system fueled with HEU if the technology can be fully recaptured and sufficient data are available to identify failure modes and benchmark modern M&S codes used to design the NTP system.^52,53 Additional full-scale testing would be required with whatever final fuel is selected that can meet the propellant temperature requirements. Selection of cermet and/or cercer fuel and/or a moderator block design and/or materials would increase technical risk, development time, and, consequentially, cost.

Phase 3 testing of integrated reactor and engine subsystems would entail significant investment of time and funding to construct a facility that can acquire the necessary performance data, manage the hydrogen effluent, and maintain safety under all planned test conditions and potential accident scenarios. As noted for Phase 2, testing of a number of NTP units would be necessary to fully retire risk and ensure repeatability of system fabrication and operation. This approach would minimize technical risk, but there would be cost and schedule risk associated with definition and construction of new facilities and obtaining environmental approvals. If only component, subassembly, and Phase 1 and Phase 2 system tests are performed along with extensive M&S validation, then a sequence of extensively instrumented flight tests, at full scale, could replace Phase 3. These flight tests could incorporate the cargo missions planned before first flight of crew. These missions would need to be carefully defined and instrumented to fully characterize system performance, including engine operation for the total throughput of LH₂ required for the baseline mission (i.e., a round-trip crewed mission). Additionally, one or more cargo missions would need to precede the planned crewed missions by a significant timeframe to allow for potential modification of the NTP system design based on data collected during the flight tests.

Without significant ground-based testing, benchmarking of M&S tools would be limited to component/subsystem hardware testing and legacy data on steady-state and dynamic performance, coolant flow, and thermomechanical and reactivity behavior from the Rover/NERVA ground testing as well as in-space testing during the cargo missions prior to the first crewed mission. A detailed review and evaluation will be required to determine the relevance of the M&S validation using Rover/NERVA test data for any new reactor design(s) and materials.

___________________

⁵¹ C. Reese, D. Burns, and J. Werner, Cost and Schedule Estimates for Establishing a Zero Power Critical Testing Capability at Idaho National Laboratory to Support NASA Nuclear Thermal Propulsion Design Development, INL/EXT-19-53988, Idaho National Engineering Laboratory, Idaho Falls, ID, 2019.

⁵² Legacy systems were fueled with HEU, and the utility of legacy research and development would be of diminished for an NTP system fueled with HALEU.

⁵³ S.K. Borowski, R.J. Sefcik, J.E. Fittje, D.R. McCurdy, A.L. Qualls, B.G. Schnitzler, J. Werner, A. Weitzberg, and C.R. Joyner, Affordable Development and Demonstration of a Small Nuclear Thermal Rocket (NTR) and Stage: How Small is Big Enough?, NASA/TM- 2016-219402, AIAA-2015-4524, 2016, https://doi.org/10.2514/6.2015-4524.

M&S predictive capabilities have advanced significantly since previous NTP development programs. The status of these capabilities to address and adequately predict coupled multi-physics simulation for NTP (to ensure startup controllability and concurrent coolability) is required before deciding on the testing path going forward. In this case, flight testing, which could be incorporated into the initial cargo missions, would become the key tool for system-level validation of successful performance, operability, controllability, coolability, reliability, lifetime, and safety. To enable this approach, the decision to freeze the flight hardware design would have to be made earlier in the development schedule, and the cargo missions would have to be launched significantly earlier than the planned 2039 crewed mission. Any technical issues identified with the first flight system would require attention, perhaps involving redesign, retest, and subsequent flight validation.

In summary, the most robust technology development program would follow testing through Phase 3, with flight tests occurring as a stand-alone mission rather than as part of a cargo mission. The lowest cost and highest risk technology development program would be to only conduct separate effects testing coupled with ZPC (Phase 1, above) and utilize flight tests conducted as part of the initial cargo missions. In this scenario, failure of the NTP on the initial cargo mission would result in significant mission delays and cost increases to support redesign and retest. There exists a spectrum of options between these two extremes. For example, ground testing through the described Phase 2 offers a lower, intermediate technical risk path, but with schedule and cost risks primarily associated with the construction of new facilities, modifications to existing facilities, safety and environmental approvals, and the completion of the testing. The schedule and cost risks associated with facilities are particularly acute for the larger facilities required for full-scale tests, especially for those that involve containment, storage, and disposal of radioactive gases, liquids, and equipment. Environmental standards, for example, are much more stringent and the environmental approval process takes much longer to complete than when full-scale test facilities were constructed for the Rover/NERVA programs. In any approach that uses the precursor cargo missions as the means for relevant-scale spaceflight demonstration, sufficient time between the first flight and the crewed missions is required to make and validate design updates.

DEVELOPMENT AND DEMONSTRATION ROADMAP

The roadmap in Figure 2.4 shows key milestones and when they would need to be achieved to execute the baseline mission: launching a crewed mission to Mars in 2039 preceded by an initial cargo mission in 2033.

The development of an NTP system for the cargo and crewed elements of the baseline mission will require several program phases. As shown in the roadmap (Figure 2.4), there is no time for delay. These phases include the following:

• Development of technology and M&S capabilities for the NTP system and its subsystems and components,
• Ground testing of subsystems and components,
• Facility development and integrated testing of the NTP system,
• Development and launch of cargo missions, and
• Development and launch of the baseline mission for human exploration of Mars.

To meet the necessary prototype demonstration schedule, several activities would need to run concurrently, including fuel architecture technology development, reactor core design, cryogenic fluid management, integrated propulsion system design, and engine component technology development and testing. Candidate fuel architectures must be evaluated to enable selection of an architecture that can meet mission requirements. The first major milestone (by the end of 2021) will be a decision to use either HEU or HALEU fuel. NASA and DOE can then initiate a fuel technology development effort to include fuel chemistry determination (UN, UCO, UO₂, etc.) and the fuel architecture technology maturation (cercer, cermet, ceramic, etc.). As shown in Figure 2.3, the successful demonstration of fuel performance will necessarily take place prior to the initiation of the prototype final design review as this fuel architecture outcome will drive final design decisions, including the choice of moderator block configuration and reactor core materials.

Technology development for reactor core structural and moderator materials are scheduled to commence in 2022 to support preliminary design efforts as facilities within industry, NASA, and DOE are available to test and validate nuclear and nonnuclear component performance and material characterization. NASA will also need to demonstrate long-term storage technologies with near-zero boiloff for LH₂ propellant tanks in the 2025 timeframe.

Engine performance considerations and the resulting reactor core operational and safety margins associated with hydrogen flow through the system will be characterized during prototype development to support the eventual demonstration of the propulsion system. It is projected that a successful prototype demonstration could be completed in the 2027 to 2029 timeframe. This will be a critical milestone in the development of the Mars flight system, the design of which must begin in 2029 or 2030 to maintain the timeline for a crewed mission in 2039.

The recommended testing regime will require new and upgraded facilities. These could become schedule-limiting without early action to develop necessary testing capabilities. Ground tests could continue into the cargo mission design phase.

Multiple cargo precursor missions are planned to deliver supplies to Mars prior to the first crewed mission, and these cargo missions could satisfy flight qualification requirements of the integrated NTP engine system. The first of these missions will need to be launched no later than 2033 to provide enough time to address any emergent issues before the 2039 crewed mission. An NTP system or the crewed mission would likely consist of multiple, largely independent, engine modules. The cargo missions may use a single NTP engine module and a lower total propellant load than is needed for the human exploration mission, but it would demonstrate the maximum propellant throughput for a single engine, and it would include enough of the performance capabilities to demonstrate adequate engine performance, lifetime, and reliability for the crewed mission.

SUMMARY

The Rover/NERVA program demonstrated the feasibility of graphite-based HEU fuels in NTP engines by ground testing nearly two dozen reactors at full power, some integrated with NTP engine hardware. Unfortunately, much of this expertise has been lost in the intervening 50 years since the program’s termination, and several design issues remain unresolved. There have been several decades of research into NTP fuels and system design since Rover/NERVA, but no NTP reactors or engines have been constructed since then, and none have ever flown. All

NASA and DOE NTP programs prior to 2013 focused on HEU designs and experiments.^54,55 In addition, only limited fuel development and M&S has been devoted to HALEU designs.

NTP development faces four major challenges that, with adequate resources, can be overcome to execute the baseline mission in 2039. As noted above, these challenges are (1) heating propellant to approximately 2700 K at the reactor exit for the duration of each burn, (2) the long-term storage of LH₂ in space with minimal loss, (3) the lack of adequate ground-based test facilities, and (4) rapidly bringing an NTP system to full operating temperature (preferably in 1 min or less).

There are currently no facilities in the United States that could conduct a full-power ground test of a full-scale NTP reactor comparable to the Rover/NERVA experiments. Existing facilities could be modified to support ZPC and low-power critical testing of an NTP reactor to validate control system status and operability, reactor excess reactivity, and shutdown margin prior to launch. NTP development can be conducted at increasing levels of complexity, starting with component testing and M&S development. Development may proceed to fully integrated reactor tests, such as ZPC tests, to verify criticality characteristics. Rover/NERVA-like experiments could be replicated to test the performance of fully integrated reactors during startup, extended operation at full power, shutdown, and restart.

The nonlinearity and scale dependence of many of the physics and potential failure mechanisms indicate the need for testing of the reactor and all tightly coupled subsystems at full scale. This may be possible through ground-based testing. Subscale NTP flight testing cannot replace full-scale ground testing. Flight qualification requirements could be satisfied by leveraging the sequence of cargo missions occurring before the first crewed mission, with the first cargo mission in the 2033 timeframe. This approach would provide sufficient time for incorporation of lessons learned into subsequent NTP cargo missions and ultimately the crewed mission in 2039.

___________________

⁵⁴ P. Venneri and Y. Kim, “Physics Study of Nuclear Reactors for Space and Rocket Propulsion,” Proceedings of the 2013 International Congress on Advances in Nuclear Power Plants, Paper KA148, 2013, http://www.proceedings.com/21535.html.

⁵⁵ G. Rosairem, et al., Design of a Low-Enriched Nuclear Thermal Rocket, Center for Space Nuclear Research, Idaho Falls, ID, 2013.