Recent Publications

Get Publication Alerts

New fusion research delivered to your inbox.

Sep 1

Alpha particle losses due to ripple and field errors in a SPARC like tokamak

Hugo Emilio Ferrari, Agustín Mairotta, Cesar F Clauser, Ricardo Farengo, Justo Andres Gonzalez Litardo, Alexandra LeViness, Steve Scott

Plasma Physics and Controlled FusionSep 1, 2026Plasma & Confinement Magnets & Superconductors AI, Modeling & Simulation

In this work we study the alpha particle losses due to ripple and coil misalignments in the SPARC tokamak. To calculate the effects of ripple and coil misalignment, we developed a code named B3D. The ripple field calculated by this code is combined with 2D equilibrium fields and predicted plasma profiles and used in the FOCUS code to calculate the alpha particle losses. We benchmarked our results against previous studies with the ASCOT and SPIRAL codes, obtaining excellent agreement. When the effect of coil misalignment is included, the alpha particle losses increase and localized regions with significantly larger values of the power density through the last closed flux surface appear. This appears to be due to the low order modes induced in the magnetic field by the misalignment.

Commonwealth Fusion SystemsMassachusetts Institute of Technology

Aug 25

Physics Attention Transformer Surrogate for Rapid Vertical Instability Growth Rate Prediction: Alcator C-Mod to SPARC

Arunav Kumar, Cesar Clauser, Theodore Golfinopoulos, Cristina Rea, Francesco Capersene, Dan Boyer, SPARC Team, Alcator C-Mod Team

arXiv (physics.plasm-ph)Aug 25, 2026Plasma & Confinement Control & Diagnostics AI, Modeling & Simulation

In this work, we investigate rapid prediction of the dominant $n{=}0$ vertical instability growth rate in C-Mod and SPARC equilibria, where nonrigid free boundary response models are too slow for control cycle use. Using a Physics Attention Transformer trained on MEQ-FGE-L labels, we predict both the scalar growth rate and the associated two dimensional perturbed toroidal current density. We find mean absolute errors of 5.4~s$^{-1}$ on held out C-Mod equilibria and 12.7~s$^{-1}$ on synthetic SPARC cases, with spatial eigenfunction errors near 5\%. We also compared PAT with operator based ML models : FNO2D and DeepONet, where we found PAT predicts a much lower normalised growth rate error and improved spatial reconstruction. These results indicate that PAT can reproduce MEQ-FGE-L outputs at control relevant latency and could support future studies of growth rate headroom monitoring and proximity aware shape control.

Aug 21

First step toward multi machine ELM energy scalings and extrapolations to SPARC and ITER

R. Perillo, A. Redl, T. Eich, C. J. Lasnier, A. Nelson, R. Rizkallah, D. Silvagni, A. Stagni, J. A. Boedo, A. McLean, et al.

arXiv (physics.plasm-ph)Aug 21, 2026Plasma & Confinement AI, Modeling & Simulation

It is shown that the ELM energy loss normalized by the plasma stored energy (ΔEELM/Wplasma) for high-density small/QCE ELM regimes scales inversely with the separatrix turbulence parameter a_t. In contrast, the neoclassical electron collisionality at the pedestal top, nu*e,neo, expected to regulate ΔEELM/Wplasma according to the Loarte scaling (Plasma Phys. Control. Fusion 2003 45 1549), does not adequately capture ΔEELM/Wplasma data for peeling-ballooning-limited type-I ELMs and ballooning-limited small/QCE ELMs, limiting its applicability for extrapolation to one scenario window. A multi-machine database including seven tokamaks and with ΔEELM/Wplasma ranging from 0.5% to 14%, has been analyzed. A regression analysis on only type-I ELMs yields ((ΔE_ELM)/W_plasma )_(Type-I ) [%]=6.8*T_(e,ped)^0.03 n_(e,ped)^(-0.4) \k{appa}^(-0.4) R_major^0.4, corresponding to ΔEELM/Wplasma =4.5% for nominal SPARC pedestal parameters and 12% for the ITER D-T Q=10 scenario. For the small/QCE ELM class, however, as a_t increases, the pedestal moves toward a ballooning-limited boundary, the toroidal mode number increases, the ELM frequency rises following the scaling f_ELM=46e^((2.25*a_t)), and ΔEELM/Wplasma decreases via the relation (ΔE_ELM)/W_plasma [%]=1.6e^(-(α_t/2)). For SPARC QCE-relevant a_t=0.86 and ITER high-fueling scenario a_t = 0.64, the scaling favorably predicts ΔEELM/Wplasma of 1.0% and 1.2%, respectively, with values below 1% if the small/QCE ELM regime is pushed beyond a_t >1. The small/QCE ELM-fitted results represent an initial step toward future analysis on broader datasets, which will be necessary to improve the accuracy of projections for future reactor-relevant scenarios.

Aug 11

Machine learning aided neutron yield for dud detection based on JET and TFTR Deuterium-Tritium plasmas

Lidia Piron, Alessandro Pau, Nicolò Ferron, Eric Fredrickson, Olivier Sauter, Matteo Baruzzo, Clive D Challis, Remi Dumont, Dirk Van Eester, Michael Fitzgerald, et al.

Nuclear FusionAug 11, 2026Plasma & Confinement Control & Diagnostics AI, Modeling & Simulation

as it indicates fusion performance. To optimize Tritium consumption and limit neutron activation, a support function included in the plasma control system called a dud detector will trigger an alarm if the plasma fails to achieve expected fusion performance. This function has been developed and routinely employed at JET during DT campaigns. This study presents machine-learning methods based on a surrogate model of the neutron rate, which can be used as an advanced dud detector. In preparation for DT operations in BEST, HL-3, ITER, and SPARC, we investigate the portability and inherent limitations of these ML methods by analysing similar DT experiments conducted at TFTR.

Max Planck Institute for Plasma PhysicsConsorzio RFXSwiss Federal Institute of Technology Lausanne

Aug 7

Characterization of a prototype parallel-plate $^{238}$U fission chamber with DD and DT fusion neutron sources

V. Hagenlocker, R. A. Tinguely, X. Wang, J. L. Ball, B. Buschmann, M. Gatu-Johnson, R. Gocht, P. Raj

arXiv (physics.plasm-ph)Aug 7, 2026Control & Diagnostics Blankets & Neutronics AI, Modeling & Simulation

The SPARC tokamak will employ $^{238}$U-based fission chambers (FCs) to monitor high-performance deuterium-tritium (DT) plasma operations, spanning neutron yield rates from ${\sim}10^{15}$ to ${>}10^{19}$ n/s. This work validates the $^{238}$U FC design, which utilizes a parallel-plate detector geometry and borated polyethylene collimation to prioritize unscattered DD and DT fusion neutrons. Experimental testing with both DD and DT neutron generators corroborates vendor-specified efficiencies and demonstrates excellent detector linearity, with measured count rates showing good agreement with OpenMC neutronics simulations. Further characterization confirms the $^{238}$U FC's robustness against SPARC-relevant environmental challenges, including stray magnetic fields up to 14 mT and possible signal degradation risks associated with $\sim$30 m long cable runs. These results confirm that the $^{238}$U FC, supported by indirect neutron shielding as well as appropriate pulse height thresholds, provides a reliable solution for fusion power measurements as part of the SPARC neutron diagnostics suite.

Deuterium retention of boron powder from deuterium gas or ion exposure

Shota Abe, Adam Q. Kuang, Christopher P Chrobak, Alessandro Bortolon, Camilo Jaramillo-Correa, Bruce E Koel

Nuclear FusionAug 7, 2026Materials & Plasma-Facing Components Tritium & Fuel Cycle

We report retention of deuterium (D), a proxy for tritium (T), from D2 neutral gas and D+ ion exposures on commercial boron (B) powder, a proxy for B dust potentially formed in fusion reactors. D removal behavior from B powder is reported as a means of estimating a potential T inventory in B dust in advanced fusion reactors, such as SPARC and ITER, the latter currently plans to employ B wall conditioning by glow-discharge boronization. B wall conditioning forms surface coatings of chemical compounds on tungsten (W) plasma-facing components (PFCs), thus suppressing plasma contamination by impurities such as oxygen, carbon, and tungsten. However, B-based slag or dust particles are expected to form. Such B dust particles can retain hydrogen isotope species, thereby causing a T inventory issue. In this work, the commercial B powder was exposed to D2 neutral gas or D+ ions and analyzed by temperature-programmed desorption to quantify D retention and the desorption temperature. The experiment confirmed D retention from D2 neutral gas exposure. A strong D2 desorption peak at 700 K, corresponding to B-D bonding, was observed for all D2 gas and D+ ion exposure cases. For D2 gas exposures, D retention was significantly enhanced at a B powder temperature of 550 K. B powder bakeout under vacuum at 600 K for 22 hours after D2 exposure exhibited efficient D removal. In contrast, the B powder bakeout at 420 K, even for 1 day, did not remove D efficiently. D retention yields from D+ ion exposure, emulating charge-exchange involving D atoms, were determined. Experiments confirmed that D retention was suppressed by oxidation of the B powder surface, which naturally occurs in reactor environments.

Princeton Plasma Physics LaboratoryCommonwealth Fusion SystemsPrinceton University

Jul 20

The power exhaust constrained SPARC separatrix operational space

B. Lomanowski, T. Eich, J. D. Lore, J. -S. Park, T. Body, P. Stangeby

arXiv (physics.plasm-ph)Jul 20, 2026Plasma & Confinement

In this work we extrapolate the separatrix operational space (SepOS) projections to SPARC and introduce detachment access criteria, thus formulating the combined power exhaust constrained SepOS (i.e., PE-SepOS) to evaluate integrated power exhaust solutions at scale. Through the interpretation of SPARC SOLPS-ITER datasets and foundational work already demonstrated in experiments we formulate an as simple as possible description of the SOL net power and momentum losses in dissipative regimes to link the main power exhaust quantities with the SepOS parameters. Through this framework, we demonstrate the utility of a normalized PE-SepOS framework in identifying accessible operational points for given exhaustible Psep requirements. In applying the PE-SepOS framework to project the SPARC operational space, we find inherent trade-offs, namely: i) accessing high impurity radiation scenarios leads to pronounced reductions in ne,sep (e.g., 50% reductions at 2% Ne concentration) as a consequence of power limitation, and ii) given present understanding of access criteria to the quasi-continuous exhaust regime (QCE), a compromise between high radiative fraction and high density/neutral pressure is required for QCE access at sufficiently high density, high alpha_t conditions, with the divertor dissipative regime transitioning to pronounced detachment. Taking advantage of universal trends enabling projections of density and impurity seeding scans, the PE-SepOS thus provides a framework for mapping out the edge plasma operational space in a scalable manner, subject to validation during early SPARC operations with suitable divertor and edge plasma observables.