arXiv subjects

Francesco Romano

Publications and source records attributed to Francesco Romano.

At least 19 recordsLinked to original sources

arXiv2026

RF Helicon Plasma Thruster for an Atmosphere-Breathing Electric Propulsion System (ABEP)

This dissertation focuses on Atmosphere-Breathing Electric Propulsion (ABEP) for propellant-less continuous orbiting in very low Earth orbits (VLEO) and very low Mars orbits (VLMO). ABEP utilizes an intake to collect atmospheric particles as propellant for an electric thruster, offsetting aerodynamic drag. An analytical system-level analysis assesses collectible mass flow, drag, and power requirements, using GOCE as a case study. Three intake designs are developed based on gas-surface interactions. Two diffuse-reflection designs yield collection efficiencies $\eta_c < 0.5$, while a specular-reflection design achieves $\eta_c < 0.95$. The specular intake demonstrates superior robustness against flow misalignment, maintaining high efficiency at large angles. The proposed ABEP thruster features a contactless, neutralizer-free design using a helicon plasma discharge, preventing erosion from aggressive species like atomic oxygen. Numerical simulations via the HELIC code determined key design parameters (frequency, magnetic field, plasma density), leading to the implementation of a resonant birdcage antenna with high electrical efficiency. The thruster was manufactured and experimentally validated using Ar, N_2, and O_2. Tests demonstrated successful ignition and operation across relevant mass flows with low power consumption ($P_f < 60$ W). Additionally, a B-dot probe was developed to detect helicon waves in the plasma plume. Ultimately, this work delivers verified intake designs and a novel contactless plasma thruster with high electrical efficiency (>99% in vacuum). Both components represent promising technologies for near-future ABEP applications. Future development outlooks, diagnostics, and spacecraft configurations are also discussed.

physics.plasm-ph↗

arXiv2026

The Diocotron Instability in the Trapped Electrons Experiment T-REX and its Relevance to Electron Clouds in Gyrotron Guns

Gyrotrons are essential for electron cyclotron resonance heating (ECRH) in fusion reactors, making their efficient operation crucial for fusion energy. Past experiments revealed instability issues due to trapped electrons in the magnetron injection gun (MIG) region, causing undesired currents and operational failures. To address this, tight manufacturing tolerances are required for the MIG geometry~\cite{pago2}. We present findings of the TRapped Electrons eXperiment (T-REX) at the Swiss Plasma Center, designed to understand electron cloud physics in gyrotron MIGs. T-REX replicates MIG geometries, electric and magnetic fields, and is supported by the 3D FENNECS code. The setup includes two coaxial electrodes in a vacuum chamber atop a superconducting magnet; a central electrode is biased to negative DC voltages and an outer one is grounded, creating a radial electric field up to 2 MV/m and an axial magnetic field B < 0.31 T. Initial discrepancies between experiments and simulations were linked to the diocotron instability, leading to FENNECS being upgraded to 3D and a dedicated set of diagnostics for T-REX. This instability causes the electron cloud to collapse and reform at a frequency depending on plasma conditions. Within this article, time-resolved current measurements on the outer electrode and top flange are presented. Further, a fast current probe array installed at the top flange is detailed. Measurements highlight rotating structures in the electron cloud resulting from the diocotron instability. Simulations show remarkable agreement with experiments, especially regarding the cloud's build-up/collapse frequency, and the rotation frequency and direction of the modes. These results improve our understanding of non-neutral plasmas in environments mimicking a real gyrotron MIG, paving the way for better gyrotron reliability.

physics.plasm-ph↗

arXiv2026

A combined dose and microdosimetric modeling framework incorporating volume effects correlates with tissue sparing in proton minibeam radiotherapy

Proton minibeam (pMB) radiotherapy, delivers highly heterogeneous dose distributions alternating high-dose peaks and low-dose valleys. This aims to widen the therapeutic window by improving normal tissue sparing while maintaining the same or even better tumour control. The performance of pMB strongly depends on the collimator design and physical parameters. To better understand the physical and radiobiological drivers of this enhanced therapeutic window, we perform a detailed microdosimetric characterization of proton minibeams and assess their impact. We characterize radiation quality with microdosimetry through Monte Carlo simulations. Then we extend the Generalized Stochastic Microdosimetric Model to predict the normal tissue complication probability (NTCP) at different depths in water, 1cm, 2cm, and 4cm, for 100MeV proton minibeams realized with varying configurations of collimator. Results are compared with conventional homogeneous field (HF) irradiation after dose normalization to the tumor. The developed model is applied by considering tissues as divided into several functional subunits, connected by a seriality parameter. Microdosimetric characterization of proton minibeam irradiation shows differences between peak and valley regions in shaping lineal energy spectra, especially at low depth, while radiation quality uniforms progressively getting closer to the tumor. NTCP calculations results suggest an increased sparing effect for pMB over conventional HF. A strong dependence is found on the peak-to-valley dose ratio (PVDR), and on the seriality parameter. Predictions indicate substantial sparing from pMB, especially for PVDR > 15, including relatively serial organs with seriality around 0.7. This integrated dose-microdosimetric-biological framework elucidates how spatial fractionation, radiation quality, and organ architecture collectively shape tissue sparing in pMB.

physics.med-ph↗

arXiv2025

Proc4Gem: Foundation models for physical agency through procedural generation

In robot learning, it is common to either ignore the environment semantics, focusing on tasks like whole-body control which only require reasoning about robot-environment contacts, or conversely to ignore contact dynamics, focusing on grounding high-level movement in vision and language. In this work, we show that advances in generative modeling, photorealistic rendering, and procedural generation allow us to tackle tasks requiring both. By generating contact-rich trajectories with accurate physics in semantically-diverse simulations, we can distill behaviors into large multimodal models that directly transfer to the real world: a system we call Proc4Gem. Specifically, we show that a foundation model, Gemini, fine-tuned on only simulation data, can be instructed in language to control a quadruped robot to push an object with its body to unseen targets in unseen real-world environments. Our real-world results demonstrate the promise of using simulation to imbue foundation models with physical agency. Videos can be found at our website: https://sites.google.com/view/proc4gem

cs.RO↗

arXiv2024

Design and First Tests of the Trapped Electrons Experiment T-REX

Gyrotrons are essential for electron cyclotron resonance heating (ECRH) in fusion reactors, making efficient operation crucial for advancing fusion energy. Past experiments revealed instability issues due to trapped electrons in the magnetron injection gun (MIG) region, causing undesired currents and operational failures. To address this, tight manufacturing tolerances are required for the MIG geometry [1]. We present initial findings of the TRapped Electrons eXperiment (T-REX) developed at the Swiss Plasma Center, designed to understand the physics of electron clouds in gyrotron MIGs. T-REX replicates MIG geometries, as well as their typical electric and magnetic fields, and it is supported by 2D Particle-in-Cell (PIC) simulations with the FENNECS code [2, 3]. The setup includes two coaxial electrodes in a vacuum chamber atop a superconducting magnet, with a central electrode biased to negative DC voltages and an outer one at ground, creating a radial electric field (1 to 2 MV/M) and an axial magnetic field (B < 0.4 T). This setup mimics Penning-Malmberg traps. We present the experimental device and first findings on current distribution and also qualitative comparison with FENNECS simulations [4]. Planned diagnostics include optical emission spectroscopy, phosphor screen imaging, Streak camera imaging, and potentially electric field distribution via the Stark effect. This research aims to enhance gyrotron performance and reliability in fusion energy systems.

physics.plasm-ph↗

arXiv2024

Performance Test Methodology for Atmosphere-Breathing Electric Propulsion Intakes in an Atomic Oxygen Facility

The testing of atmosphere-breathing electric propulsion intakes is an important step in the development of functional propulsion systems which provide sustained drag compensation in very low Earth orbits. To make satellite operations more sustainable, it is necessary to develop new materials which withstand erosion, long-lasting propulsion systems to overcome drag, and tools that allow for ground-based testing. Among the tools to enable these innovations is the Rarefied Orbital Aerodynamics Research facility at the University of Manchester. Here, a description of the facility is provided together with two different methodologies for testing sub-scaled intake designs for atmosphere-breathing electric propulsion systems. The first methodology is based on measurements of the pressure difference between the two extremities of the intake, while the second uses a gas sensor to measure the collection efficiency of the intake. Direct Simulation Monte Carlo models have been used to assess the viability of the proposed testing methodologies. The results of this analysis indicate that either methodology or a combination of both can provide suitable measurements to assess the performance of future intake designs.

physics.space-ph↗

arXiv2024

Real-World Fluid Directed Rigid Body Control via Deep Reinforcement Learning

Recent advances in real-world applications of reinforcement learning (RL) have relied on the ability to accurately simulate systems at scale. However, domains such as fluid dynamical systems exhibit complex dynamic phenomena that are hard to simulate at high integration rates, limiting the direct application of modern deep RL algorithms to often expensive or safety critical hardware. In this work, we introduce "Box o Flows", a novel benchtop experimental control system for systematically evaluating RL algorithms in dynamic real-world scenarios. We describe the key components of the Box o Flows, and through a series of experiments demonstrate how state-of-the-art model-free RL algorithms can synthesize a variety of complex behaviors via simple reward specifications. Furthermore, we explore the role of offline RL in data-efficient hypothesis testing by reusing past experiences. We believe that the insights gained from this preliminary study and the availability of systems like the Box o Flows support the way forward for developing systematic RL algorithms that can be generally applied to complex, dynamical systems. Supplementary material and videos of experiments are available at https://sites.google.com/view/box-o-flows/home.

cs.RO↗

arXiv2023

RIPTIDE: a novel recoil-proton track imaging detector for fast neutrons

Neutron detectors are an essential tool for the development of many research fields, as nuclear, particle and astroparticle physics as well as radiotherapy and radiation safety. Since neutrons cannot directly ionize, their detection is only possible via nuclear reactions. Consequently, neutron-based experimental techniques are related to the detection of charged particle or electromagnetic radiation originating from neutron-induced reactions. The study of fast neutrons is often based on the neutron-proton elastic scattering reaction. In this case, the ionization induced by the recoil protons in a hydrogenous material constitutes the basic information for the design and development of neutron detectors. Although experimental techniques have continuously improved and refined, so far, proton-recoil track imaging is still weak in laboratory rate environments because of the extremely small detection efficiency. To address this deficiency, we propose a novel recoil-proton track imaging system in which the light deriving from a fast scintillation signal is used to perform a complete reconstruction in space and time of the event. In particular, we report the idea of RIPTIDE (RecoIl Proton Track Imaging DEtector): an innovative system which combines a plastic scintillator coupled to imaging devices, based on CMOS technology, or Micro Channel Plate sensors. The proposed apparatus aims at providing neutron spectrometry capability by stereoscopically imaging the recoil-protons tracks, correlating the spatial information with the time information.

physics.ins-det↗

arXiv2023

Barkour: Benchmarking Animal-level Agility with Quadruped Robots

Animals have evolved various agile locomotion strategies, such as sprinting, leaping, and jumping. There is a growing interest in developing legged robots that move like their biological counterparts and show various agile skills to navigate complex environments quickly. Despite the interest, the field lacks systematic benchmarks to measure the performance of control policies and hardware in agility. We introduce the Barkour benchmark, an obstacle course to quantify agility for legged robots. Inspired by dog agility competitions, it consists of diverse obstacles and a time based scoring mechanism. This encourages researchers to develop controllers that not only move fast, but do so in a controllable and versatile way. To set strong baselines, we present two methods for tackling the benchmark. In the first approach, we train specialist locomotion skills using on-policy reinforcement learning methods and combine them with a high-level navigation controller. In the second approach, we distill the specialist skills into a Transformer-based generalist locomotion policy, named Locomotion-Transformer, that can handle various terrains and adjust the robot's gait based on the perceived environment and robot states. Using a custom-built quadruped robot, we demonstrate that our method can complete the course at half the speed of a dog. We hope that our work represents a step towards creating controllers that enable robots to reach animal-level agility.

cs.RO↗

arXiv2022

Imitate and Repurpose: Learning Reusable Robot Movement Skills From Human and Animal Behaviors

We investigate the use of prior knowledge of human and animal movement to learn reusable locomotion skills for real legged robots. Our approach builds upon previous work on imitating human or dog Motion Capture (MoCap) data to learn a movement skill module. Once learned, this skill module can be reused for complex downstream tasks. Importantly, due to the prior imposed by the MoCap data, our approach does not require extensive reward engineering to produce sensible and natural looking behavior at the time of reuse. This makes it easy to create well-regularized, task-oriented controllers that are suitable for deployment on real robots. We demonstrate how our skill module can be used for imitation, and train controllable walking and ball dribbling policies for both the ANYmal quadruped and OP3 humanoid. These policies are then deployed on hardware via zero-shot simulation-to-reality transfer. Accompanying videos are available at https://bit.ly/robot-npmp.

cs.RO↗

arXiv2022

ADBSat: Verification and validation of a novel panel method for quick aerodynamic analysis of satellites

We present the validation of ADBSat, a novel implementation of the panel method including a fast pseudo-shading algorithm, that can quickly and accurately determine the forces and torques on satellites in free-molecular flow. Our main method of validation is comparing test cases between ADBSat, the current de facto standard of direct simulation Monte Carlo (DSMC), and published literature. ADBSat exhibits a significantly shorter runtime than DSMC and performs well, except where deep concavities are present in the satellite models. The shading algorithm also experiences problems when a large proportion of the satellite surface area is oriented parallel to the flow, but this can be mitigated by examining the body at small angles to this configuration (${\pm}$ 0.1°). We recommend that an error interval on ADBSat outputs of up to 3\% is adopted. Therefore, ADBSat is a suitable tool for quickly determining the aerodynamic characteristics of a wide range of satellite geometries in different environmental conditions in VLEO. It can also be used in a complementary manner to identify cases that warrant further investigation using other numerical-based methods.

physics.space-ph↗

arXiv2021

System Analysis and Test-Bed for an Atmosphere-Breathing Electric Propulsion System using an Inductive Plasma Thruster

Challenging space mission scenarios include those in low altitude orbits, where the atmosphere creates significant drag to the S/C and forces their orbit to an early decay. For drag compensation, propulsion systems are needed, requiring propellant to be carried on-board. An atmosphere-breathing electric propulsion system (ABEP) ingests the residual atmosphere particles through an intake and uses them as propellant for an electric thruster. Theoretically applicable to any planet with atmosphere, the system might allow to orbit for unlimited time without carrying propellant. A new range of altitudes for continuous operation would become accessible, enabling new scientific missions while reducing costs. Preliminary studies have shown that the collectible propellant flow for an ion thruster (in LEO) might not be enough, and that electrode erosion due to aggressive gases, such as atomic oxygen, will limit the thruster lifetime. In this paper an inductive plasma thruster (IPT) is considered for the ABEP system. The starting point is a small scale inductively heated plasma generator IPG6-S. These devices are electrodeless and have already shown high electric-to-thermal coupling efficiencies using O2 and CO2. The system analysis is integrated with IPG6-S tests to assess mean mass-specific energies of the plasma plume and estimate exhaust velocities.

physics.space-ph↗

arXiv2020

A Review of Gas-Surface Interaction Models for Orbital Aerodynamics Applications

Renewed interest in Very Low Earth Orbits (VLEO) - i.e. altitudes below 450 km - has led to an increased demand for accurate environment characterisation and aerodynamic force prediction. While the former requires knowledge of the mechanisms that drive density variations in the thermosphere, the latter also depends on the interactions between the gas-particles in the residual atmosphere and the surfaces exposed to the flow. The determination of the aerodynamic coefficients is hindered by the numerous uncertainties that characterise the physical processes occurring at the exposed surfaces. Several models have been produced over the last 60 years with the intent of combining accuracy with relatively simple implementations. In this paper the most popular models have been selected and reviewed using as discriminating factors relevance with regards to orbital aerodynamics applications and theoretical agreement with gas-beam experimental data. More sophisticated models were neglected, since their increased accuracy is generally accompanied by a substantial increase in computation times which is likely to be unsuitable for most space engineering applications. For the sake of clarity, a distinction was introduced between physical and scattering kernel theory based gas-surface interaction models. The physical model category comprises the Hard Cube model, the Soft Cube model and the Washboard model, while the scattering kernel family consists of the Maxwell model, the Nocilla-Hurlbut-Sherman model and the Cercignani-Lampis-Lord model. Limits and assets of each model have been discussed with regards to the context of this paper. Wherever possible, comments have been provided to help the reader to identify possible future challenges for gas-surface interaction science with regards to orbital aerodynamic applications.

physics.space-ph↗

arXiv2020

RF Helicon-based Inductive Plasma Thruster (IPT) Design for an Atmosphere-Breathing Electric Propulsion system (ABEP)

Challenging space missions include those at very low altitudes, where the atmosphere is source of aerodynamic drag on the spacecraft. To extend such missions lifetime, an efficient propulsion system is required. One solution is Atmosphere-Breathing Electric Propulsion (ABEP). It collects atmospheric particles to be used as propellant for an electric thruster. The system would minimize the requirement of limited propellant availability and can also be applied to any planet with atmosphere, enabling new mission at low altitude ranges for longer times. Challenging is also the presence of reactive chemical species, such as atomic oxygen in Earth orbit. Such species cause erosion of (not only) propulsion system components, i.e. acceleration grids, electrodes, and discharge channels of conventional EP systems. IRS is developing within the DISCOVERER project, an intake and a thruster for an ABEP system. The paper describes the design and implementation of the RF helicon-based inductive plasma thruster (IPT). This paper deals in particular with the design and implementation of a novel antenna called the birdcage antenna, a device well known in magnetic resonance imaging (MRI), and also lately employed for helicon-wave based plasma sources in fusion research. The IPT is based on RF electrodeless operation aided by an externally applied static magnetic field. The IPT is composed by an antenna, a discharge channel, a movable injector, and a solenoid. By changing the operational parameters along with the novel antenna design, the aim is to minimize losses in the RF circuit, and accelerate a quasi-neutral plasma plume. This is also to be aided by the formation of helicon waves within the plasma that are to improve the overall efficiency and achieve higher exhaust velocities. Finally, the designed IPT with a particular focus on the birdcage antenna design procedure is presented

physics.plasm-ph↗

arXiv2020

A Generic Synchronous Dataflow Architecture to Rapidly Prototype and Deploy Robot Controllers

The paper presents a software architecture to optimize the process of prototyping and deploying robot controllers that are synthesized using model-based design methodologies. The architecture is composed of a framework and a pipeline. Therefore, the contribution of the paper is twofold. First, we introduce an open-source actor-oriented framework that abstracts the common robotic uses of middlewares, optimizers, and simulators. Using this framework, we then present a pipeline that implements the model-based design methodology. The components of the proposed framework are generic, and they can be interfaced with any tool supporting model-based design. We demonstrate the effectiveness of the approach describing the application of the resulting synchronous dataflow architecture to the design of a balancing controller for the YARP-based humanoid robot iCub. This example exploits the interfacing with Simulink and Simulink Coder.

cs.RO↗

arXiv2020

Continuous-Discrete Reinforcement Learning for Hybrid Control in Robotics

Many real-world control problems involve both discrete decision variables - such as the choice of control modes, gear switching or digital outputs - as well as continuous decision variables - such as velocity setpoints, control gains or analogue outputs. However, when defining the corresponding optimal control or reinforcement learning problem, it is commonly approximated with fully continuous or fully discrete action spaces. These simplifications aim at tailoring the problem to a particular algorithm or solver which may only support one type of action space. Alternatively, expert heuristics are used to remove discrete actions from an otherwise continuous space. In contrast, we propose to treat hybrid problems in their 'native' form by solving them with hybrid reinforcement learning, which optimizes for discrete and continuous actions simultaneously. In our experiments, we first demonstrate that the proposed approach efficiently solves such natively hybrid reinforcement learning problems. We then show, both in simulation and on robotic hardware, the benefits of removing possibly imperfect expert-designed heuristics. Lastly, hybrid reinforcement learning encourages us to rethink problem definitions. We propose reformulating control problems, e.g. by adding meta actions, to improve exploration or reduce mechanical wear and tear.

cs.LG↗

arXiv2019

Modelling Generalized Forces with Reinforcement Learning for Sim-to-Real Transfer

Learning robotic control policies in the real world gives rise to challenges in data efficiency, safety, and controlling the initial condition of the system. On the other hand, simulations are a useful alternative as they provide an abundant source of data without the restrictions of the real world. Unfortunately, simulations often fail to accurately model complex real-world phenomena. Traditional system identification techniques are limited in expressiveness by the analytical model parameters, and usually are not sufficient to capture such phenomena. In this paper we propose a general framework for improving the analytical model by optimizing state dependent generalized forces. State dependent generalized forces are expressive enough to model constraints in the equations of motion, while maintaining a clear physical meaning and intuition. We use reinforcement learning to efficiently optimize the mapping from states to generalized forces over a discounted infinite horizon. We show that using only minutes of real world data improves the sim-to-real control policy transfer. We demonstrate the feasibility of our approach by validating it on a nonprehensile manipulation task on the Sawyer robot.

cs.RO↗

arXiv2018

A Control Architecture with Online Predictive Planning for Position and Torque Controlled Walking of Humanoid Robots

A common approach to the generation of walking patterns for humanoid robots consists in adopting a layered control architecture. This paper proposes an architecture composed of three nested control loops. The outer loop exploits a robot kinematic model to plan the footstep positions. In the mid layer, a predictive controller generates a Center of Mass trajectory according to the well-known table-cart model. Through a whole-body inverse kinematics algorithm, we can define joint references for position controlled walking. The outcomes of these two loops are then interpreted as inputs of a stack-of-task QP-based torque controller, which represents the inner loop of the presented control architecture. This resulting architecture allows the robot to walk also in torque control, guaranteeing higher level of compliance. Real world experiments have been carried on the humanoid robot iCub.

cs.RO↗