4
Weapons Sciences

INTRODUCTION

The weapons sciences competency focuses on internal, transitional, and external ballistics; launch, flight, control, and navigation of guided weapons and aerial systems;¹ and the development of novel weapon concepts. The core competencies within the weapons sciences competency include aerodynamics and control; discovery, synthesis, and formulation of energetic materials; guidance and navigation of weapon systems; gun and rocket propulsion; and modeling, simulation, and experimental characterization of energetics. The Panel on Assessment of Weapons Sciences visited the U.S. Army Combat Capabilities Development Command (DEVCOM) Army Research Laboratory (ARL) within the Aberdeen Proving Ground in Aberdeen, Maryland, on October 25–28, 2022. During this visit, the panel viewed podium and poster presentations, toured facilities, and spoke to the scientific researchers within the competency. The panel organized in advance into three teams with each team assigned primary responsibility for reporting on one or more core competencies. Still, any panel member was able to provide a written comment on any relevant issue for any core competency. Below is the assessment of this competency based on their feedback to the Army Research Laboratory Technical Assessments Board (ARLTAB).

TECHNICAL QUALITY OF THE CORE COMPETENCY RESEARCH

Aerodynamics and Control Core Competency

The aerodynamics and control core competency focuses on understanding the dynamic interaction of solid bodies (vehicles and projectiles) with air and on optimization of interactions to control motion of projectiles and aerial systems.² The optimization process here refers to a holistic approach—encompassing numerical models of the physical system (vehicle and projectile structure and aerodynamic forces), trajectory planning approaches, and optimal control algorithms to develop efficient, precise, and high-performing aerial systems and projectiles. The research activities encompassed in this core competency span a wide range of systems, from hypersonics to low-speed unmanned aircraft system (UAS) technologies and address the technology areas of fluid dynamics including aerodynamics and acoustics, aeroelasticity, and electronics. The research areas that were reviewed can sensibly be separated by motivating application to one of two groups: high-speed, typically gun-launched platforms, and low speed UAS concepts. The discussion below is similarly divided into these two areas of research, with each section highlighting achievements and advancements and opportunities and challenges.

___________________

¹ U.S. Army Combat Capabilities Development Command (DEVCOM) Army Research Laboratory (ARL), “What We Do,” https://arl.devcom.army.mil/what-we-do, accessed December 20, 2023.

² DEVCOM ARL, “Foundational Research Competencies and Core Competencies,” document for the Army Research Laboratory Technical Assessment Board, received March 30, 2022.

High Speed and Hypersonic Efforts

The presentation, “Overview of Weapons Flight and Guidance Research,” summarized the challenges that are motivating the hypersonics work, clarifying that many of the technologies being developed need to accommodate extreme environments associated with gun launch and unsteadiness in flight, and they need to be capable of penetrating the adversary’s anti-access/area denial systems. These challenges lead to research on enhanced speed, range, survivability, maneuverability, and lethality and include collective approaches, Global Positioning System (GPS) denied guidance and navigation, target identification, and target tracking. The intention is to develop technologies that can be fielded in the 2028 or later time frame. Modeling is required to provide guidance in the design of robust platforms and to access flight regimes and engagement scenarios that are not amenable to testing, including the prediction of forces and moments under realistic conditions.

The scientific quality of the research conducted in hypersonics is excellent, enhanced by close interactions with the academic community and other Department of Defense (DoD) entities. Major focal areas of fundamental work are the study of laminar to turbulence transition, vortex interactions, heat loading, and shock wave boundary layer interactions. The Army has provided a canonical test article configuration, the High-Speed Army Reference Vehicle, which is being used for measurements in Mach 6 low enthalpy, quiet test facilities at Texas A&M University and Purdue University. These university facilities enable studies of laminar to turbulent transition and represent state-of-the-art collaborative efforts. As discussed below, further work is needed for the study of vortex–shock interactions, unsteadiness effects, and shock and boundary layer interactions as well as effects associated with asymmetric platforms. For many applications, these unsteady phenomena may be more important than transition, because Army applications are often at low altitude and high Reynolds number, and platform surfaces are not smooth, so transition is immediate. Stability, on the other hand is critical. For example, there is no explanation why nonlinear roll effects are observed to reduce with increasing Mach number. Flight tests being conducted at the Aerodynamics Experimental Facility and in the Transonic Experimental Facility are especially useful and represent unique capabilities.

Fundamental computational fluid dynamics (CFD) models are being developed through a strong liaison with the University of Maryland, and comparisons of various existing model predictions are providing guidance toward model utility and uncertainty quantification. While these models have not been developed at ARL, the comparisons are being conducted with good community participation. They are using KESTREL and CFD++ software, but there was no indication if they are using Overflow or FUN3D software. Those are free and, if not already in use, could be included so their performances can be compared. The current state of the art is generally considered to be Graham Candler’s US3D code (University of Minnesota). That code could be used as a point of comparison. Additional capabilities are being provided by 6.1-supported efforts by leading university researchers. One of the presenters for “Fluid Mechanics: Nonlinear Flow Interactions for Novel Control” highlighted the wind tunnel at Florida State University that can test at a high angle of attack, a new cooperative research and development agreement with North Carolina State University, and work at Georgia Institute of Technology, among others. These efforts have been supported through the Army Research Office and are now being transitioned to closer ties with the Army Research Directorate (ARD) Core Competencies teams. However, the connections between these research efforts and the ARD Core Competency teams appeared to be not yet well established, though it should be mentioned that the briefings on the connections between these research efforts were not conducted by the principal investigators.

The personnel associated with the hypersonics efforts are well respected and clearly understand the underlying scientific questions. Their close interactions with external leading researchers affirms that depth of understanding. For example, fundamental models of hypersonic flows are being developed by the University of Maryland, as exemplified by the work presented in the poster “Basic Hypersonic Flows,” which focused on hypersonic compression corners. The effectiveness and vision of team leaders, and the very active participation of many of the research personnel in the American Institute of

Aeronautics and Astronautics and other technical society meetings is noteworthy. This involvement in the community ensures technical competence and facilitates interactions with other leaders in the field.

The associated work in control of high-speed platforms recognizes that unstable platforms are more maneuverable, but that leads to significant technical challenges associated with rapid response, control authority, and potential cross coupling of control elements. These aspects were highlighted in the presentation “Flight Control/Dynamics,” which addressed the need for reduced design cycle iteration time for controllers while prioritizing robustness to model uncertainties. The work has led to the development of the Dynamic Inversion Controller for unstable hypersonic platforms. Atmospheric conditions are unpredictable, and controls need to be capable of responding to variable wind and weather effects. The goals of the research are to extend range, identify and engage targets including moving targets, and evade enemy countermeasures. One opportunity would be to expand the Army research scope in this area to include control co-design or multidisciplinary design optimization with control as a method to further ARL objectives of range extension, enhanced target engagement, and evasion of enemy countermeasures.

The work within the presentation “Flight Control/Dynamics” focused on designing an adaptive flight controller for guided projectile applications. The control logic was based on the linearized system model. The uncertainty in the aerodynamic model of the projectile was considered as an additive perturbation term in the system model. A direct adaptive control design then was proposed to deal with the uncertainty in the system model. The design was based on a deterministic system modeling. The design was tested in a simulation environment to demonstrate its effectiveness.

Direct adaptive controllers for hypersonic vehicles have a long history with bases in the “MIT rule”³ proposed in 1960s and were used in design of controllers for early hypersonic vehicles such as the X-15. Therefore, the design approach used in this project builds on strong theoretical foundations and benefits from the existing knowledge for such controllers. However, this design is deterministic and model-based. Given the limited possibility for actual flight tests, the performance analysis of the proposed controllers is strongly tied to the fidelity of the nonlinear six degrees of freedom simulator that is being used in the study. The simulation model was not discussed in the presentation, but nevertheless it is important to validate the fidelity of the nonlinear simulator, including the associated aerodynamic models through wind tunnel and flight test data. This may necessitate the inclusion of uncertainty models to improve vehicle and projectile effectiveness.

With the advances in the computing technology, modern control techniques for uncertain systems, such as the model predictive control and data-driven control techniques, have become feasible and online real-time implementable. Stochastic approaches which inherently account for more than just the boundedness of the perturbations can result in less conservative solutions with better transient response. Such techniques often deliver better results than model-based adaptive controllers. The phrase “stochastic approaches” refers to a wide array of control techniques dealing with systems with inherent randomness. Notable examples include stochastic model predictive control, stochastic robust control, and stochastic adaptive control. These methods leverage probability theory and statistical analysis to handle uncertainties beyond the boundedness of system perturbations, thereby enabling less conservative and more optimal control strategies. The Army would benefit by further developing these approaches.

Unmanned Aircraft System Efforts

An overview of the UAS research approaches and challenges was provided, which highlighted the technology pull associated with speed, endurance, reach, and agility, and the technology push associated with developing the “art of the possible.” The research areas were divided into extreme UAS (e.g., hyper-maneuverability), intelligent UAS (i.e., artificial intelligence and autonomy), and advanced UAS technologies (e.g., advanced rotor configurations, tethered UAS, and flight configurations with

___________________

³ I.M. Mareels, B.D. Anderson, R.R. Bitmead, M. Bodson, and S.S. Sastry, 1987, “Revisiting the MIT Rule for Adaptive Control,” pp. 161–166 in Adaptive Systems in Control and Signal Processing, 1986, Oxford, UK: Pergamon Press.

reduced noise generation). Nonlinear aerodynamics interactions can be used to produce high mobility in certain situations.

The importance of the hybrid design and rotorcraft analysis (HYDRA) code is recognized. It is a good tool for preliminary design of unmanned aerial vehicles (UAVs) performance including power optimization and payload. The HYDRA code was originally developed at the University of Maryland. The principal investigator on the presentation, “UAS Conceptual Sizing and Performance Framework,” is augmenting the code to improve its capabilities and is collaborating with a researcher at Texas A&M University in that endeavor. There is also an active collaboration with ONERA, the French Aerospace Research Laboratory, to evaluate HYDRA against the NASA Design and Analysis of Rotorcraft software. While HYDRA provides a preliminary assessment of the performance, and is capable of performing some optimization, it is important to verify the design configurations with high-fidelity tools and CFD. The rotor model in HYDRA is unlikely to predict the performance accurately because it is based on an integral momentum balance. It would be interesting to compare results of the HYDRA code with SUAVE⁴ from Stanford University. The use of HYDRA for improved modeling capabilities would improve agility and maneuverability to avoid future threats. An example is countering the threat of high-power lasers to low speed UAVs. HYDRA can potentially be used to design for high maneuverability, which may be required to counter direct energy threats by minimizing laser dwell time on any component. Without such capabilities, the future role of UAVs may become very limited.

The gust performance study presented constitutes a very relevant topic because it addresses improvement of the maneuverability of UAVs to respond to changes in the angle of attack. The Actuated Recirculating Gust Generator (ARGGUS) constitutes an important tool to study UAVs dynamics in urban environment (i.e., unsteady flow). The ability to generate gusts in a controlled, repeatable way is of primary importance to the design of UAVs and their control system. The unique ability to create time scales that align more closely to real gusts is valuable for ARL. This has enabled the development and validation of a new model to predict the lift during the interaction, which uses the static lift curve to predict the lift at a given effective flow angle, with corrections to account for the rate of change of the flow angle and camber of the airfoil.⁵ This model accurately predicts the lift of a blade during gust interactions and is relatively simple to implement. The work is of high quality and importance to ARL. Future developments on the response time requirements and experimental capabilities are an exciting prospect.

The stacked rotor with variable angle between the rotors represents a new idea and is definitely interesting to explore. The problem is the uncertainty on the prediction given the shortcomings of the CFD analysis. The Viscous Vortex Particle Method is crude. The use of high-fidelity CFD that could give confidence on the obtained results is encouraged.

Development of battery pairing and other hybrid approaches for addressing different flight regimes is important. The current effort focuses on the use of separate battery technologies to address rapid surge requirements (e.g., take off and rapid maneuvers) and longer cruise times. It is noteworthy that similar hybrid approaches designed to include combined battery and non-battery powered concepts might also have impact. The choice of the engine(s) strongly depends on the mission. The optimal choice will be influenced by the desired endurance, range, and thrust. For example, turbine engines, spark-ignition engines, diesel engines, fuel cells, and batteries can be considered as a function of weight and efficiency. For short endurance, the weight becomes a handicap. However, at longer endurance, the weight of fuel dominates and efficiency becomes more important in the informal optimization process. Of course, hybrid operation with a battery and any of these chemical sources adds to the possibilities.

The shared-tethered UAS effort (presentation name: “Shared-Tether UAS Trajectory Planning”) was found to have a concerning challenge to its success. The focus of this project is on the characteristics

___________________

⁴ Stanford University, “Suave,” https://suave.stanford.edu.

⁵ C. Stutz, J. Hrynuk, and D. Bohl, 2022, “Investigation of Static Wings Interacting with Vertical Gusts of Indefinite Length at Low Reynolds Numbers,” Experiments in Fluids 63(82), https://doi.org/10.1007/s00348-022-03432-7.

of copper wire tethers that would be needed to supply sufficient power to a linear chain of five or so drones. The power supplied to the tether comes either from a ground station or a “mother” ship and allows for longer flight times and high-data-rate, secure communications. Much of the analysis was on the gauge of the copper wires and the associated weight penalty for 100-meter segments. From what was presented, the bending compliance of the wires or methods for recovering from snags and tangles or adversary actions that might be enabled appeared to have not yet been considered. Such considerations would be very important because one snag could bring down the entire fleet. If such a concept is to be pursued, optical fiber methods also could be included for consideration. Optical fibers can transmit kilowatts of power, are very lightweight, are highly flexible, and can also be used for high-data-rate communication. Optical fibers might be light enough to enable multiple fiber concepts to provide redundancy in case of single fiber loss from snags or other disruptions.

Overall Soundness of the Research

The hypersonic experimentation is state of the art in the low enthalpy arena, and the codes used are also quite advanced. Things to improve include true enthalpy testing and pushing the Mach number to higher-altitude and speed conditions that require experimentation as well as modeling of more complicated physics (e.g., rarefied gas-dynamics, thermal non-equilibrium, dissociation, ionization). Each of the tasks is producing exciting results and quantifiable progress. If successful, integrating the tasks into the coupled framework would be a first-of-its-kind demonstration of such a capability and promises to offer the ability for analysis to reduce the risk of system failure across a broad range of mission profiles. There has been significant success in leveraging the scientific quality of ARL’s research by establishing essential connections with the academic community to support experimental and computational activities. Most of the efforts have focused on lower-speed hypersonics whereby chemical effects, which constitute the trademark of hypersonics, are not present. To extend this work to the domain of higher-speed hypersonic flows that include chemical effects, it would be worthwhile to using contemporary model-reduction techniques in CFD as exemplified by the work of Charbal Farhat (Stanford University), Karen Willcox (Oden Institute for Computational Engineering and Sciences, The University of Texas at Austin), and Benjamin Peherstorfer (Courant Institute of Mathematical Sciences). Neural networks can also be used to create inexpensive surrogate models for analysis that would be helpful in reducing the computational effort in the design process.

The successes including the stacked rotor and multi-element rotor blade associated with ARL’s rotorcraft research are also noteworthy and it was pleasing to see the transition of high-quality research and the impact that this research is having. There is an opportunity; however, to improve the results with the introduction of high-fidelity CFD. For example, the work done at NASA’s Ames Research Center at the Rotorcraft center, led by Robert Strawn (army employee) has developed the code named CREATE-HELIOS (rotorcraft CFD).⁶ Some interaction with the Army group at Ames Research Center on multi rotor UAVs was reported, but there appear to be opportunities to strengthen collaboration with them, as well as academic and other external partners to leverage the scientific progress achieved by the scientific community. Also, the inclusion of recent advances in reduced order modeling and ML-based surrogate models needs to be considered. There is a clear recognition of the competence of the ARL staff and researchers and their strong understanding of the fundamental physics underlying the applications. It is also clear that the researchers are aware of the relevant science conducted elsewhere.

There are opportunities for improving modeling of the rotorcraft dynamics (especially transition from vertical take-off to horizontal flight). If they are not already in communication, the ARL research team could reach out to Ames Research Center to explore opportunities with high-fidelity CFD. Helios is the state-of-the-art code for rotorcraft dynamics, developed at Ames Research Center. However, the

___________________

⁶ Y. Delorme, S.H. Frankel, R. Jain, and R. Strawn, 2018, “High-Order Large Eddy Simulation and Immersed Boundary Method on Dynamic Meshes: Application to Rotorcraft Aerodynamics,” AIAA-2018-0599, January 7, https://arc.aiaa.org/doi/10.2514/6.2018-0599.

predictions do not always agree very well with the flight data, especially in the transition maneuver, which cannot be modeled by any current code. The best prospect is likely to be through large eddy simulation using high-order numerical methods, as exemplified by the work of Z.J. Wang (University of Kansas).

There may also be educational opportunities afforded by ARL represented by the PhD works of Visal Bhagwandin in his collaboration with Pino Martin at the University of Maryland on hypersonic compression corner modeling and David Lee in his collaboration on the aerodynamics of small UAVs with Keith Moored at Lehigh University on stability in strong wind shears.

Discovery, Synthesis, and Formulation of Energetic Materials Core Competency

Accomplishments and Advancements

The discovery, synthesis, and formulation of energetic materials core competency focuses on research to propose, create, and validate compounds that can undergo rapid release of energy on demand. It also focuses on the formulation of novel energetic materials to optimize performance, manufacturing processes, and stability for explosive and propellant applications.⁷

The overall scientific quality in this competency area was very good. The synthesis and formulation methodologies being pursued are consistent with that being pursued at leading research institutions and funding agencies. The researchers in each area were aware of the state of the art in their respective fields and broadly speaking, appeared to be working at that level with specific studies working to push the state of art forward. The core competency spanned the range of energetic material synthesis and formulation, including synthesis of new molecules and formulation of new explosive mixtures via conventional approaches. New methods of generating reactive aluminum are being explored in the plasma chemistry laboratory. Novel methods of co-crystal synthesis and high-pressure synthesis development efforts were also presented.

The core competency was largely consistent with the underlying science and research conducted elsewhere. Conventional synthesis and formulation methodologies are the core mainstay of energetic development. Co-crystallization and high-pressure synthesis are novel approaches that reflect a broad understanding of new developments in the field and are expected to provide additional pathways to the synthesis of new energetics. There is currently no reliable way to predict a priori if new co-crystal formulations retain an advantage over a simple physical mixture of the two components. New co-crystal formulations instead need to be evaluated via empirical testing, which links well with the small-scale characterization methodology under the modeling, simulation, and experimental characterization of energetics core competency.

Across the full set of chemist’s presentations, no overarching synthesis and formulation methodology was articulated, and so the overall impression for most of the areas was that progress in this competency was primarily driven by having a good knowledge of the history in the field, working to make small potential changes off previous work, and then evaluating if that approach was able to improve individual properties for a given molecule or formulation. After the presentations were completed in this area, ARL illuminated their efforts to apply ML techniques to further development strategies in this competency.

Research Portfolio Challenges and Opportunities

A few opportunities for specific projects are identified below:

___________________

⁷ DEVCOM ARL, “Foundational Research Competencies and Core Competencies,” document for the Army Research Laboratory Technical Assessment Board, received March 30, 2022.

• The approach of the project “Synthesis of High Performing Propellant and Explosive Ingredients” (presentation name on the review agenda: “Novel Synthesis”) is to synthesize nitric ester compounds that could directly replace current plasticizers in explosive formulations, thereby adding additional energy to existing munitions with a minimum of additional changes necessary. This is a very reasonable goal that promises evolutionary but nonetheless significant performance enhancement with a good chance of producing a manufacturable product. The particular synthetic routes shown are known at laboratory scale and seem capable of scale up and the target choices are sound.
• The project “Disruptive Energetics—High-Pressure Stereochemistry, Energetics, and Pressure” (presentation name on the agenda: “High Pressure”) describes fundamental research on materials in extreme pressure conditions. The high-pressure results on energetic stereoisomers seem novel and very interesting in a fundamental sense but it is not clear how this knowledge can be exploited. That is not a criticism of the work, rather it is an observation that this work could be designed to attract collaborators and competitors to move it more quickly than this single effort is likely to achieve. Further understanding will likely require the examination of many model stereoisomer pairs chosen to illuminate the molecular basis of the high-pressure results. Continuing to focus on energetic compounds may be less likely to lead to useful conclusions.
• The project “Energetic Co-crystallization of High Energy Materials” (presentation name on the agenda: “Co-Crystals” [Poster]) is a straightforward materials development project that uses high-throughput screening techniques employing energetic materials that have been already produced but are not used due to critical shortcomings. If not already considered (and ruled out) there may be an opportunity to use gradient examinations that can screen an entire range of compositions in a single scan. As it stands, this routine use of the available techniques for this project are expected to produce quick results.

Speaking more broadly, the development of new formulations with properties tailored to specific applications is a critical component of any program on energetic materials. Scientists working in this area have many molecules available to evaluate for both energetic and binder components. Similarly, they have many parameters to optimize, including performance (i.e., energy release), mechanical stability, thermal stability, aging, detonation velocity for specific applications where timing is important, flowability for additive manufacturing (AM) applications, availability of formulation ingredients, and any restrictions on waste generation for new formulation processes. These are large multidimensional parameter spaces to work with.

Several successful formulation results were highlighted by each presenter with associated publications, indicating that they are performing work of high scientific quality, with excellent qualifications, and that they are utilizing their available facilities and equipment appropriately. However, there was concern that the large parameter space may stagnate in future advances. Research directions and formulation choices on each project appeared to be primarily guided by the experience and knowledge of a given chemist. This places projects at risk due to staff attrition or involvement of early-career staff with lower levels of experience and field history.

Additionally, it was not clear what the future major formation needs were (e.g., a high-temperature stable high-energy formulation or insensitive formulation with good detonation corner turning at low temperature that is printable via AM techniques), what unknown formulation physics was required for further advancement, and if they could be addressed to meet a national need in an appropriate timeframe.

Several opportunities are thus suggested that may focus future efforts for increased impact and reduced development time. First, if not already in development, it is suggested that future efforts may benefit from a database that captures what molecules are available; what prior formulation pathways have been performed with such molecules and their result (e.g., the resulting performance, sensitivity, and stability properties of each attempt). This would serve as a tool to new researchers in the area, capture the

experience of senior researchers so that it is not lost, and allow any researcher to work beyond their own experience. Secondly, it is suggested that this database could serve as the starting point for application of ML techniques to the synthesis and formulation process, with the end goal to be able to develop a capability (ML or otherwise) that would suggest molecules and formulation paths based on desired parameters as listed above. Thirdly, it may also be possible to connect these efforts to the capabilities discussed in the modeling and simulation of energetic materials (at the molecular dynamics and mesoscale levels) to future aid in the exploration of the large parameter space computationally. It is expected that this approach would benefit from deeper connections with those colleagues in the modeling, simulation, and experimental characterization of energetics core competency working on the modeling and simulation of energetic materials and ML efforts. Such a database may also incorporate extramural data on the synthesis and formulations of energetic materials and molecules from other government agencies (e.g., DoD and DOE) and open-source literature, which could be used to support ARL’s data. A discussion of the input data requirements for each “EM” (e.g., an energetic material’s physical properties and synthesis details) needed to populate the database may also be important.

The assessment criteria asks for comment on specific areas, if found, where the research may be at a major risk of not meeting its objectives. No areas were determined to be immediately at risk of not meeting objectives for this core competency. The overall balance of the core competency research portfolio appeared effective and well balanced. The bulk of the research presented was performed at the Aberdeen Proving Ground, but it was clear that several of the researchers had significant collaborations and extramural interactions other members of the community from their publication records. It is suggested that the core competency consider developing (or if in development, articulating) a methodology to identify and address long-term strategic objectives. A mentoring program that would allow transfer of corporate knowledge from experienced chemists to early career chemists would also be beneficial.

Guidance and Navigation of Weapon Systems Core Competency

Accomplishments and Advancements

The guidance and navigation of weapon systems core competency focuses on guidance, navigation, control algorithms, electronics, and design of gun-launched or missile guidance systems, to include the evaluation of new sensing modalities (inertial, magnetic, Global Navigation Satellite Systems, vision, etc.) to improve the performance of navigation and terminal homing accuracy for weapons systems in challenged or degraded spaces.⁸

This core competency’s research scope includes both internal and extramural projects focused on autonomous, multi-agent collaborative techniques for navigation, guidance, and control; and extramural projects supporting the development of atomic clocks and optical time-transfer techniques. While some elements of work contribute to fundamental theory, the majority of the work targets essential applied research in which modern theoretical advances and technologies are applied to enhance the Army’s capabilities for precise weapons delivery in challenging environments and conditions. Many of the projects address the uniquely complex challenges of midcourse navigation and terminal guidance following a high-g weapon launch.

The research review consisted of presentations, posters, and a laboratory tour, first introduced with an overview presentation covering both flight control and guidance and navigation topics. Many of the research efforts relate to midcourse navigation, terminal, and collaborative guidance, including the development and flight-testing of a unique multicomponent experimental platform, and the development of algorithms for collaborative networking, navigation, and terminal guidance. Additional posters presented complementary work on image processing and estimation algorithms. An extensive laboratory tour gave visibility into the unique facilities and capabilities of this research area. Finally, two high-level

___________________

⁸ Ibid.

presentations (“Multi-Agent Network Control,” “Stabilize/Control Networked Nonlinear Autonomous Systems and Atomic and Molecular Physics,” and “Novel PNT in Non-Ideal Environments”) summarized each of the extramural research efforts.

The research portfolio reported in the guidance and navigation of weapon systems core competency is largely focused on application-oriented research. These projects play the critical role of transitioning technology from the state of fundamental discovery through to a technology readiness level where the work is viable to be utilized by industry in order to be transitioned into use for the Army. The majority of the current internal research activities are designed to enable various aspects of collaborative navigation, target tracking, and guidance for delivery of munitions in GPS-denied environments. This includes algorithms for each of these functions, and advancement of new hardware and software technologies required to implement them. There are two extramural efforts that constitute fundamental (6.1) research in quantum sensing and timing in non-ideal environments (presentation “Quantum Sensing and Timing in Non-ideal Environments”) and multi-agent network control (presentation “Multi-Agent Network Control: Stabilize/Control Networked Nonlinear Autonomous Systems”).

The extramural projects represent fundamental 6.1 research performed by thought leaders around the country. The quantum sensing and timing program involves 76 active single investigator grants, with an emphasis on discovery of engineered entangled quantum states and development of iodine and thorium clocks for operation in non-ideal environments. This work is leading the field in scientific quality with breakthroughs achieved in iodine clock holdover performance, entangled clock states, and optical time transfer for configurable position, navigation, and timing in the field.

Much of the ARL internal research addresses the unique challenges of Army long-range precision fires—specifically midcourse navigation and precision delivery after a high-g launch. This work has a few branches that come together in the ARL Laboratory Technology Vehicle (LTV) demonstration. These highly integrated systems demonstrate cutting-edge applied research that has advanced and combined multiple technologies to enable precision midcourse navigation and terminal guidance (presentation “Mid-course Navigation & Terminal Guidance”) in challenged situations that lack navigation assistance from GPS. The systems integrate multiple technologies, including micro-electro-mechanical system inertial measurement units (IMUs), magnetometers, pressure sensors, software-defined radios for enabling navigation through signals of opportunity, as well as networks of agents for collaborative delivery. This work is leading the state of the art of applied research and development in the field, which is not performed at the same level at other institutes or research laboratories. The recent demonstration of data-driven target recognition in a live-fire experiment is particularly impressive evidence of the quality of research in this area.

The presentation “Collaborative Delivery of Networked Intelligent Munitions” showed that ARL researchers have identified several key areas where cutting edge techniques can be applied to the specific challenges of Army missions including operation in GPS denied environments, collaboration among small, resource-limited platforms operating at high speed and high dynamics. This work is currently at a relatively early stage of development, with simulation and analysis focused on implementing effective techniques to support swarming with sparse measurements and dynamic configurations. ARL researchers are aware of the need for additional advances to effectively handle loss of agents or connectivity to the network.

In reference to the assessment criteria question, “please comment on how the competency demonstrates and reflects a broad understanding of the underlying science and research conducted elsewhere?” Researchers in the guidance and navigation of weapons systems core competency generally have a broad understanding of the field and are largely building their work on state-of-the-art research in the open literature. As evidenced by the presentation “Midcourse Navigation and Terminal Guidance” and the laboratory tours, in the areas of precision delivery concepts and experimentation, midcourse navigation, and terminal and collaborative guidance, ARL researchers have demonstrated broad understanding of the underlying research conducted elsewhere by implementing both well established and relatively new approaches IMU calibration, aiding, and ranging. They have implemented hardware and software to employ state-of-the-art signal of opportunity tracking and vision-based navigation methods.

The work on image processing and scene generation (presentation name: “Image Processing and Scene Generation”) uses neural network training (a conditional generative adversarial neural network) to map visual images to infrared (IR) images. The motivation for this effort is the availability of large optical data sets, but very few in IR. They implement a deep learning–based method to transform visible-spectrum imagery into photorealistic multi-band imagery for midcourse and terminal navigation algorithm development. This research will give the Army an understanding of how ML can be used to generate synthetic imagery to support development of field-deployable algorithms. The research approach is consistent with the state-of-the-art methods in the literature for generating synthetic imagery.

Research Portfolio Opportunities and Challenges

In terms of the soundness of the competency’s research methods and methodologies (e.g., the reasoning; the use of theory, modeling and simulation, analysis, or experimentation), at the top level, ARL has carefully identified the key challenges that need to be met to achieve the target goals and is conducting research projects at various scales to address them. The guidance and navigation projects appropriately use a combination of simulations and hardware demonstrations, beginning with the most basic functionality and stepping systematically to greater complexity and realism. This is a sound approach compatible with the need to provide these capabilities in highly challenging environments.

The staging of multiple related research efforts contributing to an integrated multi-platform demonstration is well conceived. The development of the multicomponent experimental platform and demonstration of its performance successfully reaching a moving target using image-guidance shown in a video presentation provides an established capability that forms the basis for future multi-vehicle collaborative engagements. For the three Collaborative Delivery projects (presentations “Collaborative Delivery of Networked Intelligent Munitions,” “Collaborative Delivery Algorithms,” and “Collaborative Ranging and Communications”), the high-level simulation for performing trade studies of investments in performance improvement for different system components; the network focused study, and the algorithm development look farther ahead at new capabilities to be implemented.

The presentation “Collaborative Delivery Algorithms” provides a sequence of algorithms for midcourse to terminal delivery navigation of multi-munition systems in GPS-denied environments. It includes three distinct and significant algorithmic efforts: (1) bounding IMU drift in GPS-degraded environments using inter-agent ranging and some limited GPS availability; (2) developing target tracking based on aerial camera networks that maintain tracks across multiple platforms; and (3) using reinforcement learning for closed-loop terminal guidance. The work uses state-of-the-art tools such as factor graphs in Bayesian filtering and reinforcement learning and demonstrates very promising simulation results. However, details were not provided as to the error models for the IMU or the type of inter-agent ranging measurements and their errors, so it is difficult to assess the realism of the results. Consideration of biases in radio frequency (RF) ranging and gravity modeling in IMU propagation are topics that need to be addressed. Overall, the approaches taken are consistent with established methods; however, the scope of work included in the assessment review is too large to provide a rigorous assessment of the soundness of the methods as implemented in this study. The fact that the work is tightly tied to LTV testing does provide confidence that each element will be validated as it is integrated into the hardware and real time software environment.

The poster on collaborative delivery (system) studies (poster “Collaborative Delivery Studies”) takes a very effective approach developing a modular simulation environment to produce simulation results for cooperative navigation in midcourse and terminal delivery using heterogeneous sets of munitions with and without seekers. The simulation supports the expected impact of cooperative navigation in GPS-denied environments. The observability maintenance in the absence of relative bearing measurement is noted as a challenge and the overarching observability in cooperative navigation in the absence of absolute external measurements is recognized as a challenging problem. Relative pose between the agents could also be obtained by using multiple RF ranging sensors such as ultra wideband,

as has been recently explored in the literature. Given the size of the munitions, and the need for communication links to implement the collaborative delivery, this may be an approach worth considering as an alternative to using standard angle of arrival (AoA) measurements.⁹

The poster on largest ellipsoid estimation for swarm navigation (poster “Estimation Algorithms”) correctly addresses the challenge of properly accounting for correlations between state covariances in networked agents, while minimizing inter-agent communication requirements. Having determined covariance intersection methods to be too conservative, they are investigating a representation of the largest ellipsoid that fits inside the overlap region between ellipsoids. This is a timely research topic and of great relevance to the collaborative navigation objective.¹⁰

The presentation “Collaborative Delivery—Multi-Agent Localization and Connectivity Maintenance” showed that ARL researchers have applied methods similar to multi-agent work in networked systems theory that has been done in academia for other applications for more than 15 years, however, the work at ARL is being applied to a new, more challenging scenario. In this context, there is an opportunity to update with more recent advances in networked systems theory. There is a need to pay close attention to identify the specific constraints imposed by the unique problem setting and technological barriers in ARL intended applications that prevent the use of standard solutions. In the currently reported ARL work, connectivity maintenance is decoupled from the agents’ dynamical motion and resembles routing procedures. In addition, the path to decentralization was not discussed in the presentation but represents an important and not straightforward extension of the current work. Recent literature on connectivity maintenance for mobile robots examines use of the barrier function methods, which can be of relevance to the collaborative navigation problems ARL is considering.¹¹

One area, where the soundness of the approach was unclear, is the presentation, “Collaborative Ranging and Communications” on reducing antenna calibration costs within the collaborative ranging and communications topic. In considering needs for midcourse collaborative navigation, ARL has correctly identified AoA as a powerful measurement to augment ranging. As such, accurate calibration of as-installed antenna patterns would be quite important. In the given presentation, it was unclear whether the proposed work is intended to be performed in anechoic chamber or outdoors on the actual vehicle using installed software defined radio (SDR) capability. If the latter is the case, this could be an approach for simplified calibration of each unit. However, if the technique requires installation of each vehicle separately in the anechoic chamber, incorporating this system simply to reduce the time to test in anechoic chamber may not be worthwhile.

The work presented on collaborative delivery algorithms (presentation name: “Collaborative Delivery Algorithms”) is still in the phase of algorithm development and simulation, but covers a very broad range of diverse topics. They may all be developing at an appropriate pace to support the overall research goals, but perhaps more discrete, focused efforts addressing each of the elements would be beneficial.

___________________

⁹ See A. Fishberg and J. How, 2022, “Multi-Agent Relative Pose Estimation with UWB and Constrained Communication,” arXiv: 2203.11004, and Z. Cao, R. Liu, C. Yuen, A. Athukorala, B. Kai Kiat Ng, M. Mathanraj, and U-X. Tan, 2021, “Relative Localization of Mobile Robots with Multiple Ultra-Wideband Ranging Measurements,” arXiv: 2107.08842.

¹⁰ Some relevant literature that could be helpful to this research group includes the following: B. Noack, J. Sijs, M. Reinhardt, and U.D. Hanebeck, 2017, “Decentralized Data Fusion with Inverse Covariance Intersection,” Automatica 79:35–41; S. Radtke, B. Noack, and U.D. Hanebeck, 2020, “Reconstruction of Cross-Correlations Between Heterogenous Trackers Using Deterministic Samples,” IFAC-PapersOnLine 53(2):3236–3241; M. Zarei-Jalalabadi, S.M.B. Malaek, and S.S. Kia, 2018, “A Track-to-Track Fusion Method for Tracks with Unknown Correlations,” IEEE Control Systems Letters 2(2):189–194.

¹¹ See B. Capelli, H. Fouad, G. Beltrame, and L. Sabattini, 2021, “Decentralized Connectivity Maintenance with Time Delay Using Control Barrier Functions,” pp. 1586–1592 in 2021 International Conference on Robotics and Automation (ICRA), https://doi.org/10.1109/ICRA48506.2021.9561066; P. Ong, B. Capelli, L. Sabattini, and J. Cortés, 2021, “Network Connectivity Maintenance via Nonsmooth Control Barrier Functions,” pp. 4786–4791 in 2021 60th IEEE Conference on Decision and Control (CDC), https://doi.org/10.1109/CDC45484.2021.9683512.

The extramural research to stabilize and control networked nonlinear autonomous systems (presentation name: “Multi-Agent Network Control: Stabilize/Control Networked Nonlinear Autonomous Systems”) pursues innovative stochastic, data-driven, learning-based strategies that are state-of-the art in the field and are inspired by various computer science modeling tools. This research direction can lead to substantial advancement in control of autonomous systems once computational hurdles for real time implementation are overcome. The presentation on this topic focused on a single agent control problem and did not discuss multi-agent implementations, multi-agent challenges in standard environments, whether the challenges with data-driven methods are more complex than those faced for standard multi-agent control, and what innovations can these learning-based tools can bring in solving multi-agent problems. It may still be early in this project development, but an increased focus on the multi-agent challenges would be valuable to achievement of the project goals. In data-driven control, data generation can also be a significant challenge. A relevant work in the literature that looks at control using limited information is from Talebi and colleagues (2020).¹²

The project described in the presentation “Collaborative Ranging and Communications,” may be at risk of not meeting its objectives. The presentation presents two key goals: (1) develop low-cost relative localization technologies to support collaborative guidance research and (2) justify RF AoA as a viable localization technology by reducing antenna calibration cost. The presentation only discussed the antenna calibration effort, so it is not known if the first objective has been achieved. From the material presented it is not clear that the proposed methods will have a significant impact on reducing cost so as to make the AoA viable. Calibration of the antennas for AoA would likely have to be performed with the antenna mounted on the vehicle (or mockup) in the test chamber, unless the researchers are planning to implement the proposed calibration method in the field, which does not appear to be the case from the materials provided. The time required to install the vehicle in the chamber was not quantified; so, it is unclear whether the time savings for the new calibration procedures would offer a significant cost savings. The presentation also did not relate the performance of the new method to standard calibrations in terms of how they impact the intended application to localization. A final concern is that the analysis is entirely based on a high-fidelity simulation tool (HFSS). While this software is quite well established, it is understood that experimental testing is also necessary to validate an antenna design. The presentation does not mention the potential that the simulations from the HFSS might not be representative of the actual calibration performance of the proposed multi-probe methods. Development of effective AoA measurements for collaborative navigation is an important goal that would support the overall core competency. An alternative or supplementary research effort to address that topic more directly would be warranted.

In reference to the assessment criteria question on the overall balance of the core competency’s research portfolio, the research within the guidance and navigation of weapons systems core competency as presented during the review essentially includes the development and testing of laboratory technology vehicles that are able to advance, integrate, and demonstrate key technologies under the unique requirements for long range distributed and collaborative engagements; with parallel development of multiple technologies and algorithms to enable midcourse navigation and terminal guidance in contested environments. Expertise in exterior ballistics and prior success in precision delivery using GPS-guidance has positioned ARL to lead the field on alterative navigation and guidance for smart munitions.

The competence of the internal programs was clearly communicated in the presentation, “Midcourse Navigation and Terminal Guidance” as well as in the “Multicomponent Experimental Platform” poster and guidance and control laboratory tour. The posters “Image Processing and Scene Generation,” “Largest Ellipsoid Estimation for Swarm Navigation” (referred to in the agenda as “Estimation Algorithms”), and the “Collaborative Delivery (System Level) Studies” also gave details of component systems that are important for the success of the LTV program. This overall effort constitutes

___________________

¹² S. Talebi, S. Alemzadeh, N. Rahimi, and M. Mesbahi, 2020, “Online Regulation of Unstable Linear Systems from a Single Trajectory,” IEEE Conference on Decision and Control (CDC), https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9304196.

the primary research strength for the core competency in meeting the cumulative core competency goals. Building a strong connection between the advanced studies on networking and collaborative methods and the LTV projects will further solidify continued the success of the overall research portfolio.

Partnerships and collaborations with other institutes listed in the summary presentation on this work include the Armaments Center, Picatinny Arsenal, New Jersey, for electro-optical/infrared sensing, simulation, and algorithms for terminal precision; the U.S. Army Combat Capabilities Development Command Aviation and Missile Center in Huntsville, Alabama, for scene generation, data collection, and collaborative engagements; the Naval Surface Warfare Center in Dahlgren, Virginia, for alternative RF localization technologies and algorithms; the Air Force Research Laboratory in Eglin, Florida, for image based navigation technologies and algorithms; and the Defense Research and Development Canada. ARL researchers are benefiting from these collaborations and advancing the state of the art established elsewhere by addressing shortcomings when applying these techniques to low size, weight, and power (SWaP) platforms subject to high-g launches and operating in challenged environments.

Extramural academic partners working on precision network control and optimization of multi-agent autonomous systems were highlighted during the review. This program supports basic research in the stabilization and control of networked nonlinear autonomous systems; however, not enough information was received to evaluate the contribution of this work to the overall competency.

While it is expected that timing capabilities will be an important component across much of ARL’s work, the timing part of the portfolio as presented was entirely extramural. It is assumed that frequency control work is being done at ARL across other competency areas outside of weapons sciences. Timing may be an area that is fragmented across various programs. The network timing project including the iodine clocks and optical time transfer are important programs of study that present useful technologies to incorporate into internal work being done at ARL. The review presentations did not show whether this work was currently being done.

The work on collaborative navigation of networked vehicles (presentation names: “Collaborative Delivery Algorithms and Collaborative Delivery of Networked Intelligent Munitions”) brought to mind a possible opportunity to extend the work into optically networked vehicles for high-precision navigation in denied environments, and perhaps this could be another future area of research that could benefit from the work that is already happening. These networked configurable vehicles could be used to create precision navigation capabilities on an as-needed basis in challenged environments. Some of the extramural work on atomic and molecular physics—specifically efforts on optical time transfer—could be integrated with the work on networked vehicles to support this effort.

Gun and Rocket Propulsion Core Competency

Accomplishments and Advancements

The gun and rocket propulsion core competency explores the art of the possible to substantially increase range and velocity, while radically reducing the weight and form factor of future weapon systems (all calibers) and armaments for next-generation manned or unmanned (small to large air and ground) platforms.¹³ This core competency area presented two studies on AM of energetics and manufactured grains (presentation names: “Additive Manufacturing of Energetics” and “Advanced Manufactured Grains”). The scientific quality within this core competency area appeared to be good, but preliminary and primarily focused on addressing technical challenges associated with proper operation of the AM machines with energetic materials to achieve parts of desired tolerances and on concept development of AM that would enhance mission capability.

The core competency demonstrated a broad understanding of the underlying science and research conducted elsewhere by reviewing current openly published and domestic restricted capabilities. AM is a

___________________

¹³ DEVCOM ARL, “Foundational Research Competencies and Core Competencies,” document for the Army Research Laboratory Technical Assessment Board, received March 30, 2022.

new and very active field with a number of open publications with inert materials that can then be adapted to printed energetic materials. Much of the methods presented appeared to rely on exploratory research or prior publication due to the developing nature of the field. These approaches are consistent with that of other large research institutions in this area.

Research Portfolio Challenges and Opportunities

A number of challenges were articulated during the two presentations in this core competency. The researchers are currently optimizing their printing processes by working to improve the flowability of printable energetic formulations, the printed deposition uniformity and density, and the repeatability and tolerances of printed parts. Simultaneously, they are working to design and fabricate advanced grain structures with novel and improved function. These are very diverse technical challenges and the researchers may benefit from connections with other core competencies.

In answer to the assessment criteria question, this core competency isn’t at risk of not meeting its objectives but as mentioned, would likely benefit from improved connections with the other core competencies. Improved connections with the discovery, synthesis, and formulation of energetic materials core competency may assist in further development of printable formulations with improved properties and with the development of new materials specifically for AM applications. Connections with the modeling, simulation, and experimental characterization of energetics core competency may also assist more efficient development (and initial prediction) of novel grain geometries for improved capability. Via different modeling capabilities, it may also be possible to model the mechanical response of the printed material to assist in better design of the tool path and operation for improved tolerances. For example, can features such as solid particle distribution though the printed propellant, tip drag, and mechanical sagging prior to curing be predicted to reduce the number of printing “experiments” needed to generate parts of adequate mechanical properties? The issues addressed here couple strongly with expertise found in the discovery, synthesis, and formulation of energetic materials core competency. Thereby, collaboration with that group is vital.

The overall balance of the core competency research portfolio appeared effective and well balanced. The bulk of the research presented was performed at the Aberdeen Proving Ground, but it is clear from their publication records that many of the researchers had significant collaborations and extramural interactions with the community.

Modeling, Simulation, and Experimental Characterization of Energetics Core Competency

Accomplishments and Advancements

The modeling, simulation, and experimental characterization of energetics core competency focuses on the delivery of exploitable knowledge through the creation and application of physics-based theoretical models to design advanced energetic materials for increased lethality and extended range munitions and to predict their performance in system concepts.¹⁴

The overall scientific quality in this core competency area was very high and is considered cutting edge with respect to what is being done in other leading research institutions and funding agencies. The researchers in each area were aware of the state of the art in their respective fields and actively trying to advance the state of the art by applying novel techniques. This core competency was very broad and efforts encompassed both simulation development, modeling, and also experimental measurements of energetic materials. Modeling efforts spanned the range from multi-scale energetic material modeling, to the application of ML, to energetic material modeling and data analysis, and to classical continuum scale computational fluid dynamics modeling. Similarly, the experimental areas of

___________________

¹⁴ Ibid.

the core competency also covered the full range of length scales for energetic materials from the atomistic scale to the mesoscale, and up to full scale charges.

The overall balance of the core competency research portfolio appeared effective and well balanced. The bulk of the research presented was performed at the Aberdeen Proving Ground, but it was clear that many of the researchers had significant collaborations and extramural interactions with the community from their publication records.

The core competency explored promising and novel areas of research. Modeling and simulation were well represented in multiple areas. The multi-scale modeling and simulation approach constitutes a grand challenge level effort that has the potential for significant long-term impact. The ML effort is currently generating results and will likely further enhance the physical understanding of the core competency in the medium time frame. The continuum-scale CFD capability provides immediate predictive capability on difficult and computationally expensive applications. The experimental areas of the core competency were well distributed across the available range of length scales with demonstrated capability for energetic material characterization from the atomistic scale to the mesoscale, up to full scale charges. These broad ranges of supported capability in the core competency gives confidence that it will be possible to provide experimental validation data for many of the simulation scales being undertaken.

A few notable projects include:

• The project “Microscale Experiments to for Developing Novel Energetic Materials” exhibits all the desired characteristics of a successful program. The project focus appears to be of considerable value to the military. While it is not especially novel, it is entirely appropriate and is conducted efficiently—the output is very well done. The publications are in a suitable variety of forms, from technical notes that document current work to peer-reviewed publications that demonstrate scientific progress. Notably, this team is clearly at the forefront of the field and has taken leadership positions to drive the general area forward. Presentations at National meetings provide constructive feedback and identify potential collaborators. Their efforts to design symposia ensure that topics of importance to the project continue to receive attention. Overall, this work is comparable to high-quality work at universities or national laboratories.
• The project “Modeling and Simulation of Energetic Material” is an ambitious effort to model the role of microstructure in the formation and evolution of hot spots in energetic materials. This is an important question for practical munitions and the problem is well described. The effort involves many collaborators and there appears to be reasonable output. It is difficult to estimate the impact of the program at this stage from the presented material, but the overall impression is that it is competent and capable.

Research Portfolio Challenges and Opportunities

All of the presented research in this core competency was working to advance the field. When pushing the technology capability forward in a research area, it is natural to leave unexplored opportunities that need to be addressed to fully characterize the advancements. Several studies in the core competency could benefit from increased attention to two main areas. The first is increased application of error analysis and uncertainty quantification to each calculation or measurement. The second is improved linkage between model prediction and experimental measurements to ensure that new approaches are validated against an established technique. Specific examples are detailed below, along with other more specific opportunities where appropriate.

The presentation “Modeling and Simulation of Energetic Materials” focused on modeling energetic material behavior at multiple scales, referred to as the atomistic scale, the mesoscale, and the continuum scale. Atomistic or molecular dynamics calculation methods are capable of modeling the evolution of many atoms over hundreds of nanoseconds. Mesoscale simulations resolve the individual

constituent explosive components (e.g., explosive crystals, binder, and voids) at scales of up to a few millimeters, thus capturing the response of tens to hundreds of these elements. Continuum models homogenize the explosive microstructure and use continuum-level methods to predict the bulk response of the material. Currently, none of these modeling scales overlap, which presents challenges to relating any observed behavior across scales. In the present work, the atomistic scale is used to predict material parameters (i.e., heat capacity, melt curves, and thermal conductivity). Those material parameters are used as calibration parameters for a mesoscale level model, Generalized Dissipative Particle Dynamics. Data from these atomistic and mesoscale models are then used to set the parameters for a continuum scale model via a hierarchical multi-scale method, which allows an application-specific subject-matter expert to apply constraints to the continuum model, such that its bulk behavior response matches that predicted by the mesoscale.

For both of the presentations “Modeling and Simulation of Energetic Materials” and “Reaction Mechanisms/Solid State” (where the study goal was to develop a methodology to model the decomposition of energetic materials using atomistic-scale modeling and then to extend the results to continuum-level, finite-rate chemistry), the major challenge with the approach described is that it is fully reliant on calculation. Calculations at the smallest scales (atomistic) are used to generate the properties for each larger scale calculation. While this is fully self-consistent from a computational perspective, it does not connect to any experimental truth or reality. And so, there is the possibility that any errors made will both propagate and be magnified though the increasing cascade of scales in the modeling methodology. This would limit the range of validity of continuum models derived from this approach. Conventional validation of the model output at the continuum scale model will not address this issue as it may not identify which subscale parameter (or parameters) is not correct. One way to address this concern is through validation of model output at each scale to experimental measurements made at the same scale. In many cases, this is challenging or not possible (due to experimental limitations), but it is possible in some cases due to the high-experimental capability available at the local facilities or through ARL’s collaborators. In those cases, a robust assessment of uncertainty quantification (using established classical statistical techniques) may also assist in evaluation of the confidence associated with each parameter prediction.

The presentation “Machine Learning for Energetic Materials” presented as an alternative method to solve modeling challenges associated with energetic materials. Energetic materials used for engineering applications are generally composed of mixtures (formulations) of individual molecules. Each individual molecule is generally very chemically and mechanically complex. These complexities result in complex physics that currently cannot be resolved experimentally or captured by direct numerical simulation models of formulation, detonation, or mechanical loading. ML provides an opportunity to identify new correlations between material properties that are not easily identifiable via conventional methods such as inspection or application of physics-driven functional equation forms to experimental data sets. The research project has the potential to directly address the core competency goals by identifying new correlations across multiple goals. For example, ingredient combinations for formulations with targeted properties, identification of new formulations.

Several correlations that were recently identified via ML techniques were presented. They were not discussed in detail, which made it difficult to fully evaluate the overall impact of the project and, in particular, if the ML methods were identifying new relationships that were not identifiable via other means, when applied to sparse data sets, or when used with physically informed constraints versus no constraints. The overall assessment of the activity was that the work on ML is encouraging. However, an application needs to be sought when other methods do not work as well in terms of cost and accuracy. The potential of ML remains to be achieved.

For improved direction and focus the researchers could apply the ML methods to sparse data sets, with and without physics-informed constraints, and document the involved cost and accuracy against more established methods. For example, one of the topics presented identified correlations between denotation velocity and pressure. Existing partnerships (e.g., Kamlet-Jacobs) already have identified

relationships between detonation and pressure, as noted by the researchers. Thus, application of ML methods to this area is useful as a validation exercise but did not appear to identify novel relationships.

The “Dynamic X-Ray Measurements for Validating Gun Propulsion CFD Models” presentation described a study that performed experiments at the Advanced Photon Source (APS) to use X-ray imaging to validate gun propulsion CFD models by imaging the propellant combustion process in a gun cartridge and initial acceleration of a projectile in a rifle chamber. This work was clearly a preliminary effort from the recent data that was presented. The study was an excellent example of utilizing cutting-edge experimental capabilities (APS) to push the field forward. While the study was performed to validate gun propulsion models, the model data presented appeared very preliminary and possibly not yet ready for validation. This is an opportunity for future focus as well as incorporation of uncertainty quantification into the experimental results. Future experiments could also focus on better characterization of the material strain field, burn rate of propellant, obtaining more detailed resolution, and further developing the use of imaging trackers. The presented work was very important and high-quality given the challenges and limited time constraints of working at APS.

The “Dense Inert Metal Explosives” presentation described a study that was exploratory experimental research to increase detonation product energy transfer to preformed warhead casings prior to confinement failure. The current results were based on the hypothesis that an embedded pocket of metal powder in the explosive charge would result in increased case velocity prior to fragmentation. The experimental results did not support the hypothesis. This study may benefit from a more direct connection to theoretical foundation and modeling support to better explore hypotheses prior to testing. Metal powder can take a significant amount of time to react and requires excess oxygen (usually from mixing with atmosphere). And so, embedding it in an oxygen-deficient explosive may not support prompt reaction. Additionally, it may be difficult to transfer significantly more energy to cases that fragment early. The addition of light tampers (possibly reactive) between the explosive and the casing to delay case fragmentation and venting may be worth considering.

In response to the assessment criteria question asking for the identification of specific areas where the research may be at major risk of not meeting its objectives, there is the above-mentioned possibility for the multi-scale modeling effort to fail to connect to any experimental truth or reality if the results are not linked to validation experiments at each possible scale or better understood via uncertainty quantification. Otherwise, any errors might both propagate and be magnified through the increasing cascade of scales in the modeling methodology. There is also concern that the dense inert metal explosive research area may also not achieve its objectives without out better linkage to predictive modeling.

COMPETENCY PORTFOLIO OF SCIENTIFIC EXPERTISE

The scientific expertise across the weapons sciences competency was found to be very impressive. Suggestions and comments on the individual core competencies are detailed below.

ARL staff and researchers within the aerodynamics and control core competency are very competent and have a clear understanding of the fundamental physics underlying the applications. It is also clear that the researchers are aware of the relevant science conducted elsewhere.

Within the discovery, synthesis, and formulation of energetic materials, gun and rocket propulsion, and modeling, simulation, and formulation of energetic materials core competencies the qualifications of the teams of ARL researchers were found to be excellent. The teams have strong academic backgrounds and are recruiting and have personnel with a range of experience and ages. They are active in the national and international research communities in their fields as evidenced from both their collaborations and significant publication histories in technically relevant publications in their areas of research. Staff metrics showed attainment at the highest levels of research, such as fellows in professional technical societies, h-index demonstrating publications that are influential in others’ research—levels that would be equivalent to full professors in research universities. In many cases, the researchers are using and sharing tools with other laboratories and academic leaders in the field. The

approaches being used in these core competencies are of high scientific quality and are considered to be cutting edge internationally for the field of research. The approaches that are being pursued demonstrates a broad understanding of the underlying science and research conducted elsewhere.

These core competencies have the capabilities to go from the individual component research at the atomistic level to full-scale munitions. The researchers know what they need to collect, where they are trying to gain more data and insights and what equipment is needed to get the answers. Much of the equipment is developed specifically for this work. Overall, the staff is well connected to work in their areas being performed in universities and are connecting their applied work to industries that will ultimately produce the weapons systems. In only one instance was it identified that the technical approach was not well founded. These findings are consistent with staff that would be found in research universities. The portfolio reflects a small percentage reaching the pinnacle of research with awards and recognition, a core of very qualified and competent researchers at mid-career, and depending on the research area, some early-career employees with less influence and recognition.

The strong technical skill and competence of the ARL teams working on guidance, navigation, and control core competency came across clearly in several of the presentations, posters, and especially in the tour of laboratories and facilities. The research staff supporting the core competencies are qualified and eager to perform the research, evidenced by their progress on challenging research efforts and dissemination of their work at conferences and in archival publications. The successful experimental demonstration of the LTV and the progress toward a future multi-agent demonstration provides clear evidence of the strong technical skills and competence of the responsible ARL technical staff. The extramural scientists and engineers conducting research for ARL on multi-agent network control and atomic and molecular physics are among the very best in the country working on these problems and are publishing their work in the world’s top journals.

COMPETENCY FACILITIES AND RESOURCES

Suggestions and comments on how the facilities and resources are currently supporting the individual core competencies and how this might be improved are detailed below.

Overall, the facilities supporting the aerodynamics and control core competency were impressive. The flight facilities available for the study of high-speed flight are outstanding and provide unique capabilities for short-range flight testing. They include the Aerodynamics Experimental Facility and the Transonic Experimental Facility. While there was no onsite visit, their capabilities were clearly demonstrated in videos and through presented material. Utilization of NASA facilities, including the Transonic facility at NASA Langley Research Center, is also noteworthy. The ability to do IR imaging of an in-flight, high-speed platform is unique. The gust generator ARGGUS facility provides a very useful platform for testing UAS performance and developing predictive models to for rapid recovery.

For the discovery, synthesis, and formulation of energetic materials the facilities, resources, and equipment available to the researchers was excellent. The facilities appeared well organized, in very good condition, and with functional layouts. The equipment was modern and of very high quality. It did not appear that there were limitations in these areas that would impact the core competency.

The guidance navigation and control facilities and the expertise of the laboratory staff are impressive. The facility is well organized and well equipped with in-house fabrication equipment needed for rapid prototyping of printed circuit boards, custom electronics integration, an RF anechoic chamber for antenna characterization, facilities for machining, capabilities for test-firing integrated systems, test equipment for characterizing sensors, and more. These facilities have been very effectively used to develop g-hardened electronics required to support a GPS-based guidance kit and to demonstrate SDR tracking of alternative signals of opportunity. Current efforts focused on upgrades to a larger laboratory vehicle that can support positioning in GPS contested conditions using imaging and alternative RF signals of opportunity positioning are well underway. These facilities and the ongoing work here represent a key strength of the guidance and navigation of weapon systems core competency.

While there were no tours of the facilities for the gun and rocket propulsion core competency, there did not appear to be any limitations in facilities, resources, and equipment available to the researchers for this core competency.

The facilities, resources, and equipment available to the researchers within the modeling, simulation, and experimental characterization of energetics core competency are also excellent. The facilities appeared well organized, in very good condition, and with functional layouts. The equipment was modern and of very high quality. It did not appear that there were limitations in these areas that would impact the core competency. Specifically, the Crystallography Laboratory provides the ability to characterize the crystal structure of molecules and is an essential capability for an organization working in the area of energetic material synthesis and characterization. This laboratory, along with the Large-scale Indoor Performance Characterization Facility were evaluated to be excellent capabilities with significant utility to the formulation and characterization programs. The Large-scale Indoor Performance Characterization Facility contained the capability to field large charges of energetic material and characterize their detonation performance with high-speed imaging, spectroscopic analysis, and velocimetry measurements. This is also an essential capability for an organization working in the area of energetic material synthesis and characterization.

OVERALL SCIENTIFIC QUALITY OF THE COMPETENCY

Taken as a whole the scientific quality within the weapons sciences competency, along with the portfolio of the competency’s scientific expertise, facilities, and partnerships is very high. Below is a summary of salient observations for each core competency.

The work of the aerodynamics and control core competency is impressive, both in terms of scientific caliber and potential impact, especially on the experimental side. Much of the research has been published in high-quality journals. The quality benefited from close interactions with the academic community and other DoD entities. Each of the tasks is yielding exciting results and quantifiable progress. In particular, the scientific quality of the research conducted in hypersonics is excellent. The personnel are well respected, understand the underlying scientific questions, and interact well with external leading researchers. The flight tests at the Aerodynamics Experimental Facility and in the Transonic Experimental Facility are useful and represent unique capabilities. Furthermore, the CFD model comparisons are being conducted with good community participation.

The discovery, synthesis, and formulation of energetic materials core competency research quality is overall high quality with most of the work at state-of-the-art level and a few studies that are advancing that state. The researchers are aware of research activities elsewhere. Accordingly, this core competency was consistent with research quality done elsewhere by respectable teams.

The guidance, navigation, and control core competency laboratory and staff are excellent and are successfully incorporating modern techniques for navigation in GPS-contested environments and preparing for future multi-agent demonstrations that will integrate new approaches in theory, simulation, and analysis and advance guidance, navigation, and control regarding a relevant real time hardware and software demonstration on multiple vehicles.

The gun and rocket propulsion research presented was limited to AM and did not cover traditional propulsion sciences. Rather, it was closer to the energetics work in other core competencies. The work was progressing successfully and was well balanced. The researchers were aware of work done elsewhere; however, some benefits may come through stronger connection and collaboration with other core competencies. In particular, increased collaboration with experts in energetic materials in the discovery, synthesis, and formulation of energetic materials core competency and the modeling, simulation, and experimental characterization of energetics core competency would be beneficial. Improved connections with experts in energetic materials can assist in further development of printable formulations with improved properties and with the development of new materials specifically for AM

applications. Experts in energetic materials may also assist more efficient development (and initial prediction) of novel grain geometries for improved capability.

The modeling, simulation, and experimental characterization of energetics research is very high and stands at the leading edge with the best teams elsewhere (e.g., other leading institutions) in comparison. The overall research agenda has good balance and explores novel and promising areas effectively. The ML research is interesting and is encouraged, however, its potential remains to be achieved. There is good collaboration and extramural interaction although most work is done at ARL.

In general, the ARL researchers are taking advantage of emerging methodologies; the wide use of ML is an example, although still wider use is encouraged.