19th Technical Meeting on Energetic Particles in Magnetic Confinement Systems

28 Sept 2026, 08:00 → 2 Oct 2026, 18:30 Europe/Paris

Beaulieu Congress and Exhibition Center (Lausanne, Switzerland)

The 19th Technical Meeting on Energetic Particles in Magnetic Confinement Systems aims to provide a forum to discuss the status of experimental and theoretical work on suprathermal electrons and ions in a wide variety of magnetic confinement geometries, and to discuss theoretical and computational physics issues relevant to burning plasmas.

The event will focus on the formation and transportation of energetic particles, including their confinement properties. It will provide a detailed account of the location of lost energetic particle incidence on first wall components, the collective instabilities driven by energetic particles and their impact on plasma properties in different operational scenarios, and energetic particle diagnostics. It will aim, in particular, at identifying current physics issues and gaps associated with energetic particles in ITER and future fusion power plants.

 

Abstract submission and registration

• Abstract submission: please follow the Call for Abstracts link from the Menu on the left
• Registration from Local website

 

Important dates

• Local registration opens: April 2026
• INDICO abstract submission opens April 2026,
  • Abstract submission deadline (Invited, Oral): June 5th 2026
  • Abstract submission deadline (Posters): August 15th 2026
• Selection of invited and oral presentations: end of June 2026
• Local registration closes: August 15th 2026
• EP-TM Meeting: 28 September – 2 October
• ITPEA-EP Meeting: 5 – 6 October 2026

 

Publication associated with the EP Technical Meeting

As for previous editions of the meeting, selected papers from the meeting will contribute to a Focus Collection of refereed journal articles. 

The papers to be included in the collection are those selected by the Guest Editor, in consultation with the individual authors, based on criteria of originality and quality.

The Nuclear Fusion Board of Editors has kindly agreed to host the Collection. Please refer to this LOC webpage for more information about the Focus Collection.

 

Previous Meetings

Previous Technical Meetings on Energetic Particles in Magnetic Confinement Systems were held in Seville, Spain (2025), on-line (2021), Kiev, Ukraine (1989), Aspenas, Sweden (1991), Trieste, Italy (1993), Princeton, United States of America (1995), Abingdon, United Kingdom (1997), Naka, Japan (1999), Gothenburg, Sweden (2001), San Diego, United States of America (2003), Takayama, Japan (2005), Seeon-Seebruck, Germany (2007), Kiev, Ukraine (2009), Austin, United States of America (2011), Beijing, China (2013),Vienna, Austria (2015), Princeton, United States of America (2017) and Shizuoka, Japan (2019).

Monday 28 September

08:00 Doors open for badge pick-up

Welcome and Meeting opening

Convener: Mario Podesta (SPC/EPFL - Lausanne, Switzerland)

Physics of EP modes and transport

1 Experimental investigations of Alfvén eigenmode control and energetic-particle diagnostics in TCV fast-ion plasmas

Energetic particles (EPs) are essential for plasma heating and current drive in present and future fusion devices, but their confinement can be degraded by Alfvén eigenmodes (AEs) and related instabilities [1]. Developing robust methods for AE control is therefore important for preserving EP confinement and reducing fast-ion losses in reactor-relevant regimes [2]. Recent experiments on the Tokamak à Configuration Variable (TCV) have established flexible fast-ion scenarios with supra-Alfvénic neutral-beam ions, enabling dedicated studies of AE excitation, fast-ion transport, and active control.

This contribution presents recent TCV experiments demonstrating reproducible AE control using electron-cyclotron heating and current drive (ECRH/ECCD) [3,4]. Scenarios with both co- and counter-current neutral beam injection were developed to vary the fast-ion drive and access different AE regimes. In these plasmas, localized ECRH/ECCD was used to modify thermal profiles, magnetic shear, and fast-ion phase-space gradients. The experiments show that AE activity can be strongly reduced, and in some cases fully suppressed, by varying the EC deposition location and current-drive direction.

The observations indicate that AE control in TCV results from the combined action of several mechanisms. Changes in electron temperature and density profiles can modify damping, mode structure, and resonance conditions, while localized ECCD can alter the safety-factor profile and magnetic shear. AE suppression is also correlated with a reduction in fast-ion losses and redistribution, indicating a close coupling between EC actuation, wave stability, and fast-ion transport. Comparative studies in positive and negative triangularity further show that the character of AE activity is sensitive to plasma shape and equilibrium profiles, even when fast-ion confinement is similar during MHD-quiescent phases.

Reliable EP diagnostics are essential for developing such control strategies and assessing their applicability to future burning-plasma devices. In this context, recent observations of ion cyclotron emission (ICE) in TCV neutral-beam-heated plasmas provide a complementary passive diagnostic of energetic ions [5,6]. The ICE system has recently been upgraded with five additional toroidally distributed in-vessel magnetic pickup coils, enabling simultaneous multi-channel measurements and future mode-number-resolved studies of high-frequency EP-driven activity.

Together, these results demonstrate that TCV provides a flexible platform for isolating the mechanisms underlying AE control with ECRH/ECCD, while also advancing the diagnostic capability needed to characterize energetic-particle-driven modes. The findings contribute to the physics basis for active control of EP-driven instabilities and improved fast-ion confinement in future burning-plasma devices.

References
[1] W. W. Heidbrink et al., Phys. Plasmas 9, 2113–2119 (2002).
[2] M. García-Muñoz et al., Plasma Phys. Control. Fusion 61, 054007 (2019).
[3] S. E. Sharapov et al., Plasma Phys. Control. Fusion 60, 014026 (2018).
[4] A. Jansen van Vuuren et al., 30th IAEA Fusion Energy Conference, FEC2025.
[5] K. G. McClements et al., Nucl. Fusion 55, 043013 (2015).
[6] R. Ochoukov et al., EPJ Web of Conferences 346, 03009 (2026).
[7] A. Jansen van Vuuren et al., Nuclear Fusion, submitted.

Speaker: Anton Jansen van Vuuren (Swiss Plasma Center, EPFL)

2 Observations and modelling of energetic particle-driven ion cyclotron emission in MAST Upgrade

Spontaneous, strongly supra-thermal ion cyclotron emission (ICE) at or close to multiple ion cyclotron harmonics has been detected in many magnetically-confined plasmas, and is usually attributed to velocity space inversion in an energetic ion population. It thus provides a passive diagnostic of these ions that can be combined with other measurements to enhance understanding of their behaviour. Energetic particle-driven ICE in the Mega-Amp Spherical Tokamak Upgrade (MAST-U) has been detected following the addition of a high sampling rate capability to coils in an existing Mirnov array [1,2]. These measurements were carried out in pulses with either on-axis neutral beam injection (NBI) only, or a combination of on-axis and off-axis NBI. The fuel and beam ion species in MAST-U is deuterium, and ions born in the plasma at the primary injection energy are usually super-Alfvénic. ICE has been detected in MAST-U using both the high sampling rate coils and Doppler backscattering (DBS) reflectometry, thereby making it possible to obtain information on the spatial source of the emission. In one recent pulse sawtooth crashes terminated global Alfvén eigenmodes (GAEs) at 1.5 – 2.0 MHz and then excited short bursts of 2nd harmonic ICE at around 9 MHz, suggesting sawtooth-induced redistribution of fast ions driving the GAEs to the ICE-emitting region. In another case quasi-continuous, broadband ICE was excited up to the 6th harmonic, correlated with recurring instabilities in the shear Alfvén range, 100 - 300 kHz. ICE in MAST-U typically has a wide range of toroidal mode numbers n, including co- and counter-current as well as n=0, suggesting that radial gradients do not play a significant role in the instability drive. Although uncertainties remain in the ICE mode identification, preliminary linear modelling, performed by solving the local dispersion relation for compressional Alfvén waves (CAWs) in a plasma with a ring-beam distribution of energetic ions [3], show instability at frequencies and wavenumbers that are broadly consistent with the MAST-U ICE measurements.
[1] M J Hole, L C Appel, R Martin Rev. Sci. Instrum. 80, 123507 (2009)
[2] M B Dreval et al. Plasma Phys. Control. Fusion 68, 035014 (2026)
[3] J W S Cook Plasma Phys. Control. Fusion 64, 115002 (2022)

Speaker: Ken McClements (UKAEA)

3 RF modified fast-ion distribution function in the full-orbit description

Fast ions generated by neutral beam injection (NBI) or radio-frequency (RF) heating can resonantly interact with background plasma and excite Alfvénic eigenmodes (AEs), in turn leading to enhanced fast-ion transport and potential damage to reactor components [1]. The combined use of RF and NBI heating exhibits mixed effects on AE activity—mitigating or enhancing mode excitation in NSTX(-U) [2–4], and reducing AE-induced fast-ion losses in ASDEX Upgrade [5]. Understanding the RF and AE interplay with fast ions, e.g. from NBI or alpha particles in burning plasmas, is essential for the operation of future burning plasma devices.
In low-B field spherical tokamaks (STs), and particularly for high-energy RF-accelerated ions, the guiding-center (GC) approximation breaks down, necessitating a full-orbit (FO) treatment [6]. We benchmark a full-orbit RF implementation [7] against the GC quasilinear Stix operator [8] implemented in the ASCOT code [9], and validate the results against ASDEX Upgrade INPA diagnostic [10] experimental data.
The validated tool was employed to analyze NSTX(-U) cases where the High-Harmonic Fast Wave (HHFW) influences the AE stability [3]. Significant differences between FO and GC approaches are highlighted and applied to understand the heating/current drive patterns that will be possible with the new NSTX-U neutral beam system. A major difference between the approaches is RF-induced anomalous transport captured by the FO treatment. Sub- and super-diffusive regimes are identified as a function of the ratio of the resonance-layer distance to the Larmor radius, where for sufficiently small ratios the behavior returns to the diffusive limit, implied by the Stix operator. We will discuss the effects on AE stability, especially for poorly confined, highly energetic ions in low-B-field NSTX scenarios. To investigate the effects of RF waves on the AE saturation regime, a perturbative method for AE mode amplitude and phase evolution [11, 12] has been implemented in ASCOT5. We will use this tool to characterize AE properties, with emphasis on AE chirping as they are routinely observed at the NSTX(-U) tokamak, and to develop strategies for employing RF waves as an AE control mechanism in tokamaks.
[1] A. Loarte et al., Plasma Physics and Controlled Fusion 45 (2003) pp. 1549-1569
[2] W. W. Heidbrink et al., Plasma Physics and Controlled Fusion 48 (2006) pp. 1347-1372
[3] E. D. Fredrickson et al., Nuclear Fusion 55 (2015) pp. 013012
[4] M. Podestà et al., Nuclear Fusion 52 (2012) pp. 094001
[5] J. Galdon-Quiroga et al., in preparation
[6] G. Kramer et al., Plasma Phys. Control. Fusion 55 (2013) 025013
[7] P. Oyola et al., in preparation
[8] T. H. Stix et al., Nuclear Fusion 15 (1975)
[9] E. Hirvijoki, PhD Thesis, 2014
[10] J. Rueda-Rueda et al., Plasma Physics and Controlled Fusion 66 (2024) 035008
[11] R. B. White, The theory of toroidally confined plasmas.
[12] A. Bierwage et al., Plasma and Fusion Research 16 (2021) 1403087

Speaker: Pablo Oyola (Princeton Plasma Physics Laboratory)

10:35 Coffe berak

Synergy EP and turbulence

4 Theory of Zonal Fields Beat Driven by Alfvén Eigenmodes in Tokamak Plasmas

We have further extended previously derived expressions for zonal electromagnetic fields (ZFs) beat driven by reversed shear Alfvén eigenmode (RSAE) [L. Chen et al., Nucl. Fusion 65 (2025) 016018] to the case of toroidal Alfvén eigenmode (TAE). It is found that the expression of zonal current (i.e., the zonal component of the parallel vector potential) remains robust. The expression of the zonal flow (i.e., the zonal component of the scalar potential including finite poloidal-harmonics coupling), however, could be quantitatively and qualitatively modified; depending on the relative time scales between thermal ions and zonal fields, as well as the ratio between the energetic-particle (EP) and thermal-ion pressures. In particular, while the energetic-particle polarization is always neoclassical; the thermal-ion polarization could be either neoclassical (for slow zonal time scale) or classical (for fast zonal time scale). In the case of Alfvén instabilities strongly excited by EPs, as in some current experiments and simulations, one has classical polarization for the thermal ions, and the resultant zonal flow could then, for typical parameters, be an order of magnitude larger than that given in [L. Chen et al., Nucl. Fusion 65 (2025) 016018]; which was derived under assumptions of thermal-ion neoclassical polarization and negligible EP contribution . Implications of the present results to the regime of burning plasmas will be discussed.

Speaker: Prof. LIU CHEN (UNIVERSITY OF CALIFORNIA, IRVINE)

5 On the Beat Driven Nonlinear Excitation of Zero Frequency Zonal Fields by Alfvén Eigenmodes

Simulations and experiments have shown that zonal flows can play a crucial role in the regulation of turbulence. In the context of energetic particles physics for magnetic confinement devices, it is then interesting to study how the interplay of energetic particles and Alfvén eigenmodes can affect the dynamics of zonal flows, and more in general zonal structures, and potentially impact the confinement and self-organization of a burning plasma. In particular, the present work focuses on further theoretically understanding how Alfvén eigenmodes destabilized by energetic particles can generate zonal fields via non-linear mode-mode coupling. After showing that previous literature can be considered satisfying only for the case of single dominant poloidal harmonic eigenmodes (e.g., reversed shear Alfvén eigenmodes), for which the zonal flow generation is mediated exclusively by the thermal plasma, we explore the role of energetic particles in the mediation of the non-linear interaction for a generic Alfvén eigenmode (e.g., toroidal Alfén eigenmodes with two dominant poloidal harmonics). In the framework of non-linear gyrokinetic theory, the investigation is divided into two parts. On the one hand we refine previous analytical treatments in the simplified case of well-passing particles in circular cross section geometry, on the other hand we formulate a semi-analytical approach valid in general tokamak geometry for both passing and trapped particles. Comparisons with the gyrokinetic code ORB5 are also carried out. While the current work still focuses on the linear growth stage of the Alfén eigenmodes, corresponding to the beat-driven growth phase of the zonal fields, future work including the non-linear saturation regime and the impact of the investigated zonal fields on plasma confinement is foreseen.

This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200 — EUROfusion). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.

Speaker: Riccardo Stucchi

6 First Experimental Studies of Energetic Particle-Turbulence Dynamics in Spherical Tokamak Plasmas

This report presents experimental and modeling studies on the interaction between energetic particles, Alfvén eigenmodes (AEs), and thermal plasma turbulence in neutral-beam-heated MAST-U spherical tokamak plasmas. These studies address the nonlinear coupling between fast-ion-driven MHD activity and microturbulence, with emphasis on the impact of energetic particles on turbulence regulation and transport. While previous experiments on conventional aspect-ratio tokamaks have shown evidence of AE-mediated turbulence suppression [1] and nonlinear wave coupling leading to zonal-flow generation [2], this work provides the first detailed investigation of these processes in a spherical tokamak configuration, which possesses a fundamentally different turbulence spectra. Results from MAST-U discharges are presented with distinct beam-on and beam-off phases containing altered fast ion driven modes and turbulence characteristics. Preliminary analysis indicates that the turbulence and AE amplitudes are weakly anti-correlated. Doppler backscattering (DBS) provides the primary turbulence measurements making it possible to study the impact of fast ions and fast ion-driven modes on: turbulence spectra, nonlinear coupling behavior, radial, temporal, and toroidal correlations, and perpendicular flow and radial electric field profiles. Macroscopic transport quantities, such as $\chi_i$- and $\chi_e$-profiles, from integrated TRANSP [3] simulations are compared against the DBS measurements. Higher fidelity is supplied from CGYRO [4] simulations which classify the turbulent spectra and dominant modes. Lastly, experimental analysis of fast ion transport is presented using data from a fast ion loss detector (FILD), neutral particle analyzer (NPA), fast ion D-$\alpha$ (FIDA) spectrometer, and neutron diagnostics. Overall, this work establishes an experimental and computational basis for multiscale energetic particle–turbulence interactions in low-aspect-ratio plasmas and provides insights relevant to the optimization of thermal confinement and fast-ion performance in spherical tokamaks.

[1] X.D. Du et al. 2025 Phys. Rev. Lett. 135 265101
[2] J. Ruiz Ruiz et al. 2025 Phys. Rev. Lett. 134 095103
[3] A.Y. Pankin et al. 2025 Comput. Phys. Commun. 312 109611
[4] J. Candy, E. A. Belli, and R. V. Bravenec 2016 J. Comput. Phys. 324 73

Speaker: Phillip Bonofiglo (PPPL - Princeton Plasma Physics Laboratory (US))

7 Fine Structure of Sheared Flow Driven by Fishbone Instability in Tokamak Plasma

In tokamaks, it is well established that sheared flows mitigate turbulence that drives heat and particle transport, thereby providing a direct pathway to improved confinement. In particular, achieving the ITER steady-state and long-pulse high-performance operation relies on the internal transport barrier (ITB), a narrow region of steep pressure gradients that improves core plasma confinement.

Fishbone instability, one of the particular concerns in ITER hybrid scenarios, has been found to be beneficial for the formation of ITB in ASDEX Upgrade, HL-2A and EAST. One hypothesis is that fishbone-driven E×B sheared flows generated through the expulsion or redistribution of energetic particles may in turn affect turbulence. However, no experimental evidence has firmly substantiated the hypothesis due to the lack of direct flow diagnostics in the core region. Whether fishbone causes the ITB formation or not is still unclear.

In this work, we report the first direct experimental evidence of E×B sheared flows driven by fishbone instability. The flow exhibits a fine and radially reversed structure within the q=1 rational surface, rather than the global pattern reported in previous simulations [1]. Global nonlinear gyrokinetic simulations using the GTC demonstrate that the flow is driven by a zonal radial electric field generated by the instability. The simulated radial profiles agree quantitatively with Doppler reflectometry measurements [2]. Notably, the fine structure is determined by both the thermal ions and electrons rather than fast ions. The contributions from thermal ions and electrons to the radial electric field E_r are comparable in magnitude but opposite in direction. The zonal electron density response is therefore crucial, as it partially cancels the zonal ion density, reducing the net charge separation and thus significantly weakening the total amplitude of E_r.

[1] G. Brochard, C. Liu, X. Wei, et al., Saturation of fishbone instability by self-generated zonal flows in tokamak plasmas. Physical Review Letters 132, 075101(2024).
[2] Gao L, Liu A, Ding W, et al. Measurements of the E×B velocity fluctuation associated with the fishbone instability using the Doppler reflectometry on EAST[J]. Nuclear Fusion 65, 116019(2025).

Speaker: Yuehao Ma (University of Science and Technology of China)

12:50 Lunch break

Poster session I and Coffee break

8 Implementation of full-orbit energetic-particle dynamics in the kinetic–MHD hybrid simulation code MEGA

We develop a full-orbit delta-f energetic-particle model in the hybrid kinetic-MHD code MEGA and apply it to nonlinear simulations of a beam-driven, sub-cyclotron global Alfven eigenmode (GAE) in a low-field DIII-D plasma. The equilibrium and beam parameters are based on the NSTX-similarity DIII-D discharge reported by Heidbrink et al. (Nucl. Fusion 46, 324, 2006), for which the observed sub-cyclotron modes were later identified as counter-propagating GAEs by linearized HYM simulations (Belova et.al. Nucl.Fusion 62,106016, 2022). By resolving the full gyromotion and particle trajectories, the present model retains finite-Larmor-radius effects and captures the Doppler-shifted cyclotron resonance beyond the guiding-center approximation.
The most unstable counter-propagating mode occurs at n = -16 with a frequency of 0.60 times the on-axis deuteron cyclotron frequency, in close agreement with the HYM result. A beam-density scan reproduces an instability threshold of a few percent. The mode exhibits shear-Alfven polarization in the core, low poloidal harmonics, and core localization, and its frequency lies just below the shear-Alfven continuum accumulation point, identifying it as a GAE. The mixed polarization observed toward the plasma edge is consistent with the compressional signature reported in the original DIII-D experiment. Wave-particle energy-transfer analysis identifies co-passing energetic ions as the driving population through the fundamental Doppler-shifted cyclotron resonance.

Nonlinear evolution reveals that resonant energetic particles are redistributed along resonance characteristics, analogous to the characteristic-constrained transport observed for conventional Alfven eigenmode-induced fast-ion redistribution. The redistribution relaxes the pitch-angle anisotropy responsible for the instability, resulting in local flattening of the resonant energetic-particle distribution at saturation. These results demonstrate that the full-orbit delta-f framework enables self-consistent investigation of cyclotron-resonant wave-particle interaction and nonlinear saturation of sub-cyclotron Alfvenic instabilities.

Speaker: Dr Hanzheng Li (National Institute for Fusion Science)

9 Observations and modelling of energetic particle-driven ion cyclotron emission in MAST Upgrade

Speaker: Ken McClements (UKAEA)

10 ORB5 simulations of the fishbone triggering conditions

For operation of future tokamak scenarios, a high pressure gradient favours the bootstrap mechanism, but also makes the plasma prone to MHD instabilities [1]. One of these instabilities is the fishbone instability, which is a 𝑚=𝑛=1 internal kink mode driven unstable by energetic particles [2]. It has been reported by both experimental and
numerical studies that the fishbone instability is not only redistributing EPs in the tokamak core but can also lead to increased ion temperature gradients via the suppression of background turbulence [3, 4]. On ASDEX Upgrade, the fishbone instability have been observed before the Internal Transport Barrier (ITB) formation [5]. In order to understand the detailed physics mechanisms and conditions for this effect, a global, kinetic non-linear approach is necessary that allows us to describe the cross-coupling between this non-perturbative global instability and the electromagnetic micro-scale fluctuations.

In this poster, simulations with the gyrokinetic particle-in-cell code ORB5 [6] are performed, to first find the triggering condition for the fishbone instability. The fraction of energetic particles is scanned across three different magnetic equilibria, including reversed shear. Two energetic particle distribution functions are compared, a
Maxwellian-like distribution and a strongly anisotropic one with 𝑣 ∥ ≈ 0, which maximizes the trapped-particle fraction. For the Maxwellian distribution function, increasing the fraction of energetic particles increases the growth rate of the mode, which appears to be a toroidal Alfvén eigenmode (TAE). The anisotropic distribution function with 𝑣 ∥ ≈ 0 causes a reduction in growth rate, the TAE mode is lost, and an internal kink mode is recovered, where the growth rate decreases with EP density. The results presented show how the observed modes vary between
the scanned parameters. Further scans will investigate the triggering conditions for the fishbone instability as a preparation for comprehensive studies under experimentally relevant conditions.

This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom
Research and Training Programme (Grant Agreement No 101052200 – EUROfusion). Views and opinions expressed are however those of the
author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the
European Commission can be held responsible for them.
References
[1] Richard Lake. Consequences of Fast Ion Driven Modes in MAST. Ph.d. thesis, University of Warwick, Centre for Fusion, Space and
Astrophysics, May 2013.
[2] Liu Chen, R. B. White, and M. N. Rosenbluth. Excitation of internal kink modes by trapped energetic beam ions. Phys. Rev. Lett., 52:1122–
1125, Mar 1984.
[3] G. Brochard, C. Liu, X. Wei, W. Heidbrink, Z. Lin, N. Gorelenkov, C. Chrystal, X. Du, J. Bao, A. R. Polevoi, M. Schneider, S. H. Kim,
S. D. Pinches, P. Liu, J. H. Nicolau, and H. Lütjens. Saturation of fishbone instability by self-generated zonal flows in tokamak plasmas.
Phys. Rev. Lett., 132:075101, Feb 2024.
[4] S. D. Pinches, S. Günter, A. G. Peeters, and ASDEX Upgrade Team. Fishbone generation of sheared flows and the creation of transport
barriers. In 28th EPS Conference on Controlled Fusion and Plasma Physics, volume 25A, pages 57–60, Funchal, June 18–22 2001. European
Physical Society. ECA Vol. 25A (2001) 57-60.
[5] S. Günter, A. Gude, J. Hobirk, M. Maraschek, S. Saarelma, S. Schade, R.C. Wolf, and ASDEX Upgrade Team. Mhd phenomena in advanced
scenarios on asdex upgrade and the influence of localized electron heating and current drive. Nuclear Fusion, 41(9), 2001.
[6] E. Lanti, N. Ohana, N. Tronko, T. Hayward-Schneider, A. Bottino, B.F. McMillan, A. Mishchenko, A. Scheinberg, A. Biancalani, P. An-
gelino, S. Brunner, J. Dominski, P. Donnel, C. Gheller, R. Hatzky, A. Jocksch, S. Jolliet, Z.X. Lu, J.P. Martin Collar, I. Novikau, E. Sonnen-
drücker, T. Vernay, and L. Villard. Orb5: A global electromagnetic gyrokinetic code using the pic approach in toroidal geometry. Computer
Physics Communications, 251:107072, 2020

Speaker: Erik Eidsvig (IPP - Max-Planck-Institut für Plasmaphysik (EU))

11 Experimental Characteristics of Core-localized AEs during ICRH High-βp Plasmas on EAST

Ion cyclotron resonance heating (ICRH) is a key auxiliary heating method for fusion reactors[1], supporting multiple heating schemes including minority, harmonic, and three-ion heating. These schemes generate fast ions that can efficiently drive Alfvén eigenmodes (AEs)[2], which can induce fast-ion redistribution or loss, potentially impacting plasma-wall interactions and first-wall components.

EAST has studied high-$\beta_p$ discharges ($B_t$ = 2.5 T, $I_p$ = 400 kA), which are characterized by a high fraction of bootstrap current, excellent confinement properties, and the potential for non-inductive steady-state operation, and are considered promising candidates for future fusion reactors[3], using pure radio-frequency heating ($P_{\rm ECRH}$ = 2 MW, $P_{\rm LHW}$ = 2 MW, $P_{\rm ICRH}$ = 2–4 MW) with hydrogen minority heating (H/(H+D) $\sim 4\%$) to drive core-localized TAE (n = 3)[4] and BAE (n = 2). Investigating fast-ion-driven AEs and their impact on confinement in this regime is highly relevant for the development of reactor-relevant plasmas.

The core fast-ion beta, $\beta_f = 2\mu_0 p_f/B^2$, is identified experimentally as a key parameter for the excitation of AEs. The experimental results indicate the existence of a finite $\beta_f$ threshold for AE destabilization, with the TAE threshold higher than the BAE threshold, i.e. $\beta_{f,\mathrm{TAE}}^{\mathrm{thre}} > \beta_{f,\mathrm{BAE}}^{\mathrm{thre}}$. The AE amplitude decreases as $\beta_f$ is reduced and eventually vanishes when $\beta_f$ falls below the corresponding threshold.

Experimentally, when the fast ions have sufficient energy to resonate with TAE, they tend to preferentially destabilize TAE rather than BAE. BAE is destabilized only when the fast-ion drive is insufficient to excite TAE, suggesting a competition between TAE and BAE for the available fast-ion free energy.

The injection of NBI beams is observed to have a stabilizing effect on both TAE and BAE. Since the beam energy is relatively low, $E_{\rm NBI} \sim 55~\mathrm{keV}$, the injected beam ions may contribute more to damping than to resonant drive for these modes. In addition, second-harmonic deuterium heating can compete with hydrogen minority heating for the absorbed ICRH power. Numerical results show that the power fraction absorbed by deuterium is only about 4%, but this can still reduce the fast-hydrogen-ion drive. From this perspective, injecting relatively low-energy NBI beams to enhance damping relative to fast-ion drive may provide an effective approach for mitigating AE activity.

No clear degradation of fast-ion or bulk plasma confinement is observed for these core-localized AEs. Under the current long-pulse high-$\beta_p$ operating scenarios, these AEs appear to have little impact on steady-state plasma operation.

[1] M. Salewski et al., Nucl. Fusion 65, 043002 (2025).
[2] F. Zonca and L. Chen, Phys. Plasmas 21, 072120 (2014).
[3] S. Ding and A. M. Garofalo, Rev. Mod. Plasma Phys. 7, 4 (2023).
[4] C.Y. Pan et al., Nucl. Fusion 66, 016026 (2026).

Speaker: Dr Chenyu Pan (Institute of Plasma Physics, Hefei Institutes of Physical Science Chinese Academy of Sciences, Hefei 230031, China)

12 Energetic-particle observations of second harmonic tritium heating in JET DTE3 and comparison with H-minority heating

Energetic-particle observations of second harmonic tritium heating in JET DTE3 and comparison with H-minority heating

M.J. Mantsinen $^{1,2}$, P. Jacquet$^{3}$, D. Gallart$^{1}$, K. Kirov$^{3}$, E. Lerche$^{3,4}$, D. Taylor$^{3}$, C.D. Challis$^{3}$, E. Delabie$^{5}$, A. Kappatou$^{6}$, D. Keeling$^{3}$, D. King$^{3}$, V. Kiptily$^{3}$, M. Nocente$^{7,8}$, E. Parr$^{3}$, S. Silburn$^{3}$, E.R. Solano$^{9}$, Z. Stancar$^{3}$, E. Tsitrone$^{10}$, JET Contributors$^{11}$ and the EUROfusion Tokamak Exploitation Team$^{12}$

$^{1}$Barcelona Supercomputing Center, Barcelona, Spain
$^{2}$ICREA, Barcelona, Spain
$^{3}$United Kingdom Atomic Energy Authority, Culham Science Centre, Abingdon, Oxon, OX14 3DB, United Kingdom of Great Britain and Northern Ireland
$^{4}$Laboratory for Plasma Physics, ERM/KMS, B-1000 Brussels, Belgium
$^{5}$Oak Ridge National Laboratory, TN 37830 Oak Ridge, USA
$^{6}$Max-Planck-Institut für Plasmaphysik, Boltzmannstr. 2, 85748 Garching, Germany
$^{7}$Institute of Plasma Science and Technology, CNR, 20125 Milano, Italy
$^{8}$Dipartimento di Fisica ‘G. Occhialini’, Universit`a di Milano-Bicocca, Milano, Italy
$^{9}$Laboratorio Nacional de Fusión, CIEMAT, 28040 Madrid, Spain
$^{10}$CEA, IRFM, F-13108 St-Paul-Lez-Durance, France
$^{11}$See Maggi et al 2024 (https://doi.org/10.1088/1741-4326/ad3e16) for the JET contributors.
$^{12}$See Joffrin et al 2024 (https://doi.org/10.1088/1741-4326/ad2be4) for the EUROfusion Tokamak Exploitation Team.

Second harmonic ICRF heating of tritium is the reference ICRF heating scenario for ITER D–T plasmas. It has previously been investigated in D–T plasmas on TFTR [1-2] and JET [3-6], including high-performance JET plasmas during the second main D-T campaign (DTE2) at JET [7]. During the third main D–T campaign (DTE3) at JET, the scheme was revisited in H-mode plasmas and directly compared with hydrogen minority heating, which is an established ICRF heating scheme at JET with good heating performance.

Four DTE3 discharges, 104547–104550, were performed to investigate second harmonic tritium heating and to compare it with H-minority heating. The two heating scenarios were compared under similar plasma conditions with combined deuterium neutral beam injection and ICRF heating, allowing their heating performance and energetic-particle characteristics to be assessed. Neutral particle analyser measurements, neutron and gamma spectroscopy,lost fast ion diagnostics, and PION modelling are used to investigate the fast-ion populations and ICRF power partitioning. The two heating schemes produced similar ion temperatures, neutron rates and sawtooth-free periods, while the second harmonic tritium case resulted in lower electron temperature and stored plasma energy. Markedly different energetic-particle signatures were nevertheless observed as expected: clear energetic H and D fluxes were measured during H-minority heating, whereas no H, D or T fluxes above background were detected during second harmonic tritium heating.

The neutron spectra of the two second harmonic tritium discharges also show differences associated with plasma density, providing additional information on the underlying energetic-ion populations. PION modelling is used to investigate the role of intrinsic and the resulting partition of ICRF power between fast ions, bulk ions and electrons. The DTE3 measurements provide new experimental information on energetic-particle generation during second harmonic tritium heating and new benchmarks for modelling this ITER-relevant ICRF scenario.

References
[1] Phillips C.K. et al 1995 Phys. Plasmas 2 2427
[2] McGuire K.M. et al 1995 Phys. Plasmas 2 2176
[3] Start D.F.H. et al 1998 Phys. Rev. Lett. 80 4681
[4] Start D.F.H. et al 1999 Nucl. Fusion 39 321
[5] Eriksson L.-G. et al 1999 Nucl. Fusion 39 337
[6] Rimini F.G. et al 1999 Nucl. Fusion 39 1591
[7] Mantsinen M. J. et al 2023, Nucl. Fusion 63 112015

Speaker: Mervi Mantsinen (Barcelona Supercomputing Center (EU))

13 Influences of fast ions from neutral beam injection in synergy with ICRF waves on fishbone instability

Synergistic heating using Ion Cyclotron Range of Frequencies (ICRF) waves and Neutral Beam Injection (NBI) can produce fast ions (FIs) with energies significantly higher than those of the beam ions. The effects of NBI-ICRF synergy (NIS) FIs on NBI-fast-ion-driven fishbone instabilities are investigated using the M3D-K code. The FIs distribution function contains a slowing-down distribution and a drifting bi-Maxwellian distribution, which represent the NBI FIs and the NIS FIs, respectively.
When the equilibrium modification induced by NIS FIs is neglected, the NIS FIs exerts a stabilizing influence on the NBI-driven fishbone mode. This stabilizing effect is enhanced with the increasing perpendicular temperature $T_\perp$ and beta $\beta_{NIS}$ of the NIS FIs. A high-frequency fishbone branch is driven when $\beta_{NIS}$ exceeds an onset threshold. Besides, a fully stable region may occur for sufficiently large $T_\perp$, in which case the fishbone is suppressed before the high-frequency fishbone is driven. These results indicate a fishbone-free regime under NBI-ICRF synergy conditions. When the equilibrium modification by NIS FIs is taken into account, the net effect becomes stabilizing for larger $T_\perp$ and destabilizing for smaller $T_\perp$.
The generalized energy principle indicates that the stabilizing contribution mainly originates from the non-resonant kinetic response of the NIS FIs, which depends on $T_\perp$. In contrast, the adiabatic contribution, corresponding to the equilibrium modification, is destabilizing and is independent of $T_\perp$. Therefore, the overall impact of NIS FIs is determined by the competition between the adiabatic and kinetic effects.

Speaker: Runzhe Zhang (University of Science and Technology of China)

14 ASCOT-RFOF modelling of ion cyclotron heated fast ion loads on ASDEX Upgrade antenna limiters

The Monte Carlo orbit-following code ASCOT /1/ is a widely used tool for simulating fast ion minority populations in toroidal devices, featuring source modules for fusion products (AFSI) and neutral beam injection (BBNBI) as well as modelling of ion cyclotron resonance heating (RFOF). In addition to its capability to simulate the formation of hot ion tail distributions from various sources in realistic plasma configurations, ASCOT’s main strength is its ability to model fast ion heat load distributions on experiment-specific plasma-facing components described by sets of triangles imported from stereolithography (STL) CAD data.

In the present work, ICRH simulations are made for ASDEX Upgrade discharge #42762 using ASCOT-RFOF with a new, detailed 3D wall mesh based on STL data for the four ICRF antenna limiters and Faraday screens, the four passive limiters, and the complete set of protection tiles for the upper and the lower magnetic perturbation coils.

In addition to the new wall model, a new postprocessing scheme is applied where the STL data describing a given plasma-facing component is refined by recursively splitting large triangles into smaller ones until a requested area per triangle is achieved. The recorded exact locations of fast ion marker hits are then mapped onto the refined mesh, demonstrating that, marker hit statistics permitting, ASCOT is able to simulate fast ion heat load patterns on an arbitrarily detailed mesh of any plasma-facing component (see Fig. 1).

For details of the thermo-mechanical analysis of the results and comparison to experimental observations, see the contribution of R. Ochoukov et al. in this Technical Meeting.

Figure 1. Top: fast ion load on antenna 4. Bottom: refined mesh and load on one limiter tile.

/1/ E. Hirvijoki et al., Comput. Phys. Commun. 185 (2014) 1310.

Speaker: Seppo Sipilä (Aalto University)

15 Assessment of 1D critical gradient predictions of fast-ion transport in high q-min, high β, DIII-D steady-state scenario discharges

Several reduced models exist for predicting anomalous energetic particle (EP) transport that may be induced by instabilities such as Alfven Eigenmodes, spanning a wide spectrum of fidelity and computational cost. However, these models have typically been validated against isolated experimental discharges instead of databases, providing an incomplete assessment of their reliability and sensitivity to input uncertainty. In this poster, progress is presented on the validation of the TGLF-EP+Alpha critical gradient model [1] against a set of carefully analyzed, high q-min, high β, DIII-D steady-state scenario discharges where substantial EP transport has been observed [2,3]. TGLF-EP iteratively solves a gyro-Landau fluid system, balancing the fast-ion drive of the most unstable mode against thermal damping at each radial grid point to determine the critical EP density gradient corresponding to marginal stability. Under the assumption of stiff transport, the Alpha code relaxes the EP density profile to the critical gradient, also yielding a corresponding fast-ion diffusivity profile that can be used in TRANSP with the Monte Carlo code NUBEAM for integrated modeling and experimental validation. The ability of TGLF-EP+Alpha to reproduce the experimental observations in these plasmas is evaluated in two ways. First, TRANSP/NUBEAM is used to calculate the neutron rate due to the relaxed EP density and compare it to the measured neutron rate. Second, the NUBEAM-calculated EP distribution, including the diffusivity from Alpha, is used with FIDASIM forward modeling to compare the predicted to measured EP density profile. Moreover, the TGLF-EP+Alpha model sensitivity to equilibrium and kinetic profile input uncertainty is assessed in these discharges. Ongoing work and future plans towards large scale, quantitative validation of TGLF-EP+Alpha against a large and diverse database of DIII-D discharges will be discussed.

[1] R.E. Waltz, E.M. Bass, W.W. Heidbrink, and M.A. Van Zeeland, Nucl. Fusion 55, 123012 (2015)
[2] W.W. Heidbrink, J.R. Ferron, C.T. Holcomb, et al., Plasma Phys. Control. Fusion 56, 095030 (2014)
[3] C.T. Holcomb, W.W. Heidbrink, J.R. Ferron, et al., Phys. Plasmas 22, 055904 (2015)
*Work supported by US DOE under DE-FC02-04ER54698 and DE-SC0020337

Speaker: Jeff Lestz (University of California Irvine)

16 Behavior of kinetic instabilities in a dynamically forming resonant distribution

Instabilities driven by energetic particles are central to the physics of a burning plasma. The majority of kinetic simulations and reduced models assume that the unstable distribution is already fully established when energetic-particle-driven modes grow unstable. In realistic scenarios, however, energetic particles may accumulate in the resonance on an effective timescale comparable to the growth rate of the instability, meaning that the formation of the resonant distribution and the growth of the unstable mode must be treated concurrently. We study the behavior of these instabilities in the presence of such a dynamically forming distribution, evaluating two distinct metrics which measure how close a mode is to its linear stability threshold and how close a mode remains to its nonlinear stability threshold. It is found that saturation at large $\omega_b/\nu_\text{eff}$ (where $\omega_b$ is the bounce frequency of deeply trapped particles and $\nu_\text{eff}$ is the effective scattering rate at a resonance), normally associated with strongly driven excitation, can be achieved even if dynamically the mode remains at all times near its nonlinear stability threshold. We extend existing analytic models for near-marginal and far from marginal modes allowing for a time-dependent linear growth rate, deriving explicit expressions for the mode amplitude evolution. These formulas are shown to agree with nonlinear kinetic simulations. The discrepancies between the case of a dynamically forming distribution and the case of a fully formed distribution are shown to be particularly pronounced for energetic particle distributions which relax diffusively.

Speaker: Eamon Hartigan-O'Connor (Princeton University)

17 Effect of externally applied magnetic perturbations on ICRH fast-ion confinement in ASDEX-Upgrade

A. Reyner-Viñolas1,2, J. Galdon-Quiroga1, J.M. Ordonez-Jimenez1, J. Gonzalez-Martin3, J. Rueda-Rueda4, Ye.O. Kazakov5, R. Ochoukov6, R. Bilato6, W. Suttrop6, S. Sipilä7, L. Sanchis1, the ASDEX Upgrade teama and the WPTE teamb
1. Department of Atomic, Molecular and Nuclear Physics, University of Seville, 41012 Seville, Spain
2. CNA (U. Sevilla, CSIC, J. de Andalucia), Sevilla, Spain
3. Department of Mechanical and Manufacturing Engineering, University of Seville. Seville, Spain
4. Department of Physics and Astronomy, University of California, Irvine, CA 92697, United States of America
5. Laboratory for Plasma Physics, LPP-ERM/KMS, EUROfusion Consortium member, TECPartner, Brussels, Belgium
6. Max Planck Institute for Plasma Physics, Boltzmannstr. 2, 85748 Garching, Germany
7. Department of Applied Physics, Aalto University, PO Box 14100, 00076 AALTO, Finland
a) See author list of T. Pütterich et al, 2026 Nucl. Fusion 66 116002
b) See the author list of N. Vianello et al 2026 Nucl. Fusion 66 116010

The confinement of fast ions, generated by either the auxiliary heating systems or product of fusion reactions, is critical for future power plants, as they can severely damage plasma facing components and reduce performance. External magnetic perturbations (MPs), commonly used to control edge localized modes (ELMs) present in high confinement regimes, can either improve or degrade the overall fast-ion confinement depending on the applied MP spectrum. Previous works demonstrated that external magnetic perturbations (MPs), commonly used to control edge localized modes (ELMs) present in high confinement regimes, can either improve or degrade the overall NBI fast-ion confinement depending on the applied MP spectrum [1, 2]. In this contribution, we analyse the MP effects over NBI and ICRH fast-ion populations, which have higher energy and are more toroidally symmetric than NBI.

Experiments have been carried out in AUG, the only European tokamak equipped with both ICRH and MP capabilities. Two ICRH schemes are studied: 3rd harmonic D heating at Bt = 1.7 T, and minority-H heating at Bt = 2.5 T. In both cases > 3 MW of ICRH are applied, simultaneously to deuterium NBI and Electron Cyclotron Resonance Heating (ECRH). The main diagnostic employed in this work is the Fast Ion Loss Detector (FILD) [3], used to measure the fast-ion losses velocity-space evolution.

Simultaneous NBI and ICRH-induced losses have been observed in FILD. NBI losses originating from both the HFS and the LFS have been identified, corresponding to passing and trapped orbits respectively. The modulation of these losses corresponds to global rotation of the MPs. However, different regions of the velocity-space have a different phasing of the modulation. ICRH losses occur at higher energies (~100 keV) than NBI injection (60 keV). The modulation of the ICRH losses follows the change in the differential phasing of the MPs and it has a lower relative amplitude to the total signal measured. These results suggest that MPs can be used as actuators for phase-space engineering of higher-energy, toroidally symmetric fast-ion populations.

[1] L. Sanchis et al 2021 Nucl. Fusion 61 046006 (2021)
[2] J. Galdon-Quiroga et al 2022 Nucl. Fusion 62 096004 (2022)
[3] M. Garcia-Muñoz et al., Rev. Sci. Instrum. 80, 053503 (2009)

Speaker: Alex Reyner Viñolas (University of Seville)

18 Energetic-ion Driven Instabilities during I-phase and inter-ELM in the HL-3 Tokamak

In HL-3 H-mode plasma discharges, two new MHD instabilities have observed and identified. Both instabilities are localized in the outer region or the pedestal region of the plasma. The first is a low-frequency instability with f=0–50 kHz, exhibiting rapid frequency chirping and a toroidal mode number of n=3. It propagates in the ion diamagnetic drift direction, displays energetic-particle mode (EPM) behavior, and triggers continuous I-phase oscillations. The second is a high-frequency instability with f=100–500 kHz. The most unstable mode has a toroidal mode number of n=0. This instability family sometimes manifests as three frequency bands exhibiting an integer harmonic relationship. The most unstable mode frequently couples with an n=1 BAE mode in the outer region to generate numerous sideband modes. These instabilities tend to appear during the ELM-free phase and preceding ELMs, but can also be observed in Ohmic heating plasma, albeit with weaker amplitude. This most unstable mode exhibits global Alfvén eigenmode (GAE) characteristics.

Speaker: Wei Chen (SWIP)

19 Fast-ion transport in EAST plasmas under the combined tearing mode and edge-localized mode perturbations: a particle-tracing simulation

Fast-ion transport in EAST plasmas is studied using a particle-tracing method for two cases: edge-localized-mode (ELM) electromagnetic perturbations alone and coexisting ELM and tearing-mode (TM) perturbations. The ELM perturbations are obtained from nonlinear BOUT++ simulations first and then mapped onto the triangular equilibrium mesh used in the Particle orbit Tracing Code (PTC). When ELM perturbations alone are considered, the total fast-ion loss fraction increases only slightly, by approximately 0.7% of the total fast-ion loss compared with the unperturbed case. Nevertheless, clear fast-ion acceleration is observed in the simulations, supporting the contribution of the parallel component of the ELM-induced electric field to the acceleration process. The resulting ELM-induced transport is localized near the plasma boundary around the inner and outer midplanes, and the toroidal deposition exhibits filament-like structures. For the coexisting TM and ELM perturbations, the threshold for TM-induced orbit stochasticity remains barely changed, while the loss characteristics depend on the TM perturbation amplitude. This amplitude dependence is manifested in two regimes: near the TM-determined stochastic threshold, ELM perturbations moderately increase the total number of lost fast ions but nearly double their local deposition density; at larger TM amplitudes, the small drift-island chains generated by TM-ELM synergy, which can already appear near the stochastic threshold, become sufficiently effective to offset ELM-enhanced edge losses, leading to a reduction in the total fast-ion loss.

Speaker: Feng Wang DLUT (Dalian University of Technology (CN))

20 Investigation of mode transition induced by fast particle transport in phase space on EAST

Numerical investigation of mode transition induced by fast particle transport in phase space on the EAST tokamak has been carried out by the global kinetic-magnetohydrodynamic (MHD) code M3D-K. Firstly, based on the fishbone instabilities observed on EAST, linear simulations show that the fishbone instability is excited at experimental value of beam ion pressure. When the beam ion pressure is increased to exceed a critical value, the low frequency fishbone transits to a beta-induced Alfvén eigenmode (BAE) with much higher frequency. Nonlinear simulations show that the frequency of the low frequency fishbone chirps up and down with corresponding hole-clump structures in phase space. In addition to the low frequency mode, the high frequency BAE is excited during the nonlinear evolution due to the fast particle redistribution in phase space. Secondly, Energetic Particle Modes (EPMs) have been observed in EAST neutral beam injection (NBI) plasma, and a transition from EPMs to toroidal Alfvén eigenmodes (TAEs) is found. Nonlinear simulations show that the EPM frequency chirps down, and then a TAE emerge due to the fast particle transport in phase space. Finally, fishbone instabilities characterized by mode numbers m/n = 1/1, 2/2 have been observed in EAST. Numerical simulations show that the m/n = 2/2 high frequency fishbone branch is linearly stable, but nonlinearly grows due to the coupling with the m/n = 1/1 low frequency fishbone branch. The m/n = 2/2 fishbone frequency is almost twice of the m/n = 1/1 fishbone, and both fishbone frequencies chirp down together. In addition, a BAE with m/n = 2/2 is observed in the late nonlinear stage of the simulation. Energetic particle nonlinearity is dominant for the m/n = 2/2 fishbone saturation and transition to BAE. These works all find the mode transitions of different modes induced by fast particle transport in phase space.

Speaker: Wei Shen (Institute of Plasma Physics, Chinese Academy of Science)

21 Neoclassical collision physics regularization of 4D fast-ion distribution reconstructions in constants of motion space

Achieving burning fusion plasmas relies on the confinement of energetic fast ions, yet this remains challenging due to the gaps in our understanding of the fast-ion distribution function. However, the reconstruction of high-dimensional fast-ion distribution function is a severely ill-posed inverse problem. As a result, to find physically meaningful solutions, the experimental data must be augmented by prior information.

In this work, we present a method for reconstructing four-dimensional fast-ion distributions in JET using synthetic data. The approach employs a large set of basis functions, which are energetic particle distributions simulated by ASCOT. In these simulations, energetic ions are injected on a grid in constants-of-motion space and tracked in time obeying neoclassical collision physics. Because this domain describes all topologically allowed orbits in the plasma, it serves as a natural framework to capture all physically possible reconstructions of the fast-ion distribution function. The basis functions are evolved only over a partial slowing-down time to capture correlations between neighboring basis functions. We characterize the properties of these basis functions, analyze the resulting correlation structures, and demonstrate that the approach improves the fidelity of reconstructed four-dimensional fast-ion distributions.

Speaker: Otso Hyvärinen (University of Helsinki)

22 Numerical Investigation of NBI–ICRF Synergistic Fast-Ion Phase-Space Evolution and Neutron Signatures in EAST Plasmas

The synergy between neutral beam injection (NBI) and ion cyclotron range of frequencies (ICRF) heating provides an effective route for tailoring fast-ion distribution functions and enhancing non-thermal fusion reactivity in tokamak plasmas. In this work, NBI–ICRF synergistic fast-ion physics in an EAST plasma scenario is investigated using complementary TRANSP and PTC simulations. NBI-only and combined NBI–ICRF cases are compared in terms of fast-ion inventory, mean-energy and density profiles, energy–pitch-angle distributions, neutron production channels, and synthetic neutron spectra along selected lines of sight. TRANSP is used to evaluate fast-ion sources, NBI deposition, RF power absorption, and thermal, beam–thermal, and beam–beam neutron components, while PTC is employed to resolve the orbit-level phase-space evolution of resonant beam ions.
The simulations show that ICRF wave–particle interaction selectively accelerates resonant NBI-born ions, broadens the fast-ion energy distribution, and forms a pronounced high-energy tail. The resulting changes in pitch-angle distribution and real-space localization modify the relative contributions of different neutron production channels and lead to measurable changes in line-of-sight-resolved neutron spectra. These responses are attributed to RF-driven beam-ion acceleration, velocity-space anisotropy, and finite-orbit-width effects. Sensitivity scans of the cyclotron-resonance-layer position, antenna-launched parallel wave number, and NBI beam tangency radius further clarify how heating geometry controls RF absorption by beam ions and the associated neutron enhancement. The combined TRANSP–PTC analysis establishes a simulation framework linking beam-ion seed formation, resonance-selective RF acceleration, fast-ion phase-space reconstruction, and neutron diagnostic signatures, supporting the interpretation and optimization of NBI–ICRF synergy experiments on EAST.

Speaker: Xiangfeng Wu (Dalian University of Technology)

23 Observation of nonlinear mode coupling and fast ion loss during sawtooth oscillations in tokamak plasmas

We report the analogous experimental observation of the nonlinear mode coupling preceding sawtooth crashes on HL-2A and HL-3 tokamaks. In this experiment, nonlinear mode coupling between m/n = 1/1 internal kink mode (where m and n are poloidal and toroidal mode number respectively) and energetic particle modes (EPMs) is conclusively identified by MEGA analysis, and the mode coupling is specified as characterized as the interaction between the n = 1 kink mode and n = 2 EPM to generate n = 3 sideband. Furthermore, remarkable fast ion loss following crashes is also discussed. Numerical simulations successfully reproduce the process through a comparison of a series of cases with different mode numbers. The results show that although the initial transport of fast ions is induced by the n = 1 internal kink mode, the dramatic subsequent loss is due to the nonlinear mode coupling process. These findings demonstrate the impact of nonlinear interactions on the confinement of energetic particles (EPs). Active control strategies should therefore be developed for multi mode nonlinear coupling to mitigate the potential loss of alphas in future fusion devices.

Speaker: Yufan Qu (Southewestern Institute of Physics)

24 On nonlinear saturation of toroidal Alfvén eigenmode due to thermal plasma nonlinearities

The confinement and slowing-down of energetic particles (EPs) generated by fusion reaction and/or neutral beam injection (NBI) are essential topics in achieving self-sustained burning plasma [1]. Alfvén instabilities, which can be driven unstable by EPs [2], can induce significant EPs transport [3], among which the toroidal Alfvén eigenmode (TAE) is an important component due to its minimal continuum damping in toroidicity-induced gaps [4, 5]. Thus, the nonlinear saturation mechanism of TAE is an essential topic. Previous studies focused on the nonlinear saturation of TAE due to EPs phase space nonlinearity, which predicted the quadratic dependence of saturation level of TAE on the linear growth rate. For future tokamaks with strong EPs activities due to fusion reaction and/or neutral beam injection, the saturation level of TAE due to EPs nonlinearity can be very high.
Thus, in this work, the effects of thermal plasma nonlinearities on the saturation of TAE are investigated using both gyrokinetic particle-in-cell (PIC) code ORB5 [6] and theory. More specifically, thermal plasma nonlinearity investigated here contributes to the excitation of zonal field fluctuations as well as phase space zonal structure (PSZS) of thermal ion and electrons [7], via beating of single-$n=6$ TAE and itself. The simulations are performed in both cases with/without $n=0$ zonal fields, while PSZS always exists in PIC simulation once the corresponding particle species evolves nonlinearly.
In the simulation without zonal fields, it is found that the saturation level of TAE is dominated by thermal plasma nonlinearity for $γ_L/ω>0.47%$, which is $eδϕ_n/T_e∼0.1$ and almost independent of linear drive; While the saturation level with only EPs nonlinearity is $eδϕ_n/T_e∼1$. Upon saturation, the decrease of mode frequency and separation of $m=10$ and 11 poloidal harmonics can be observed, resulting from PSZS of thermal plasmas. Meanwhile, the saturation level can be quantitatively obtained by the gyrokinetic theory.
In the simulation with zonal fields, it is found that their introduction can significantly counteract the effects of PSZS of thermal plasmas, leading to a factor 2 enhancement to the saturation level of TAE compared to the case without zonal fields, i.e., the case with only PSZS. Meanwhile, in both cases with/without zonal fields, the saturation level of TAE is found to be proportional to the square root of inverse aspect ratio, which is also predicted by the theory [8,9]. This fact implies a stronger TAE activity in devices with larger inverse aspect ratio.

Reference:
[1] L. Chen and F. Zonca, Review of Modern Physics 88, 015008 (2016).
[2] L. Chen, Physics of Plasmas 1, 1519 (1994).
[3] A. Fasoli, C. Gormenzano, H. Berk, et al., Nuclear Fusion 47, S264 (2007).
[4] C. Cheng, L. Chen, and M. Chance, Ann. Phys. 161, 21 (1985).
[5] G. Y. Fu and J. W. Van Dam, Physics of Fluids B 1, 1949 (1989).
[6] E. Lanti, N. Ohana, N. Tronko, et al., Computer Physics Communications 251, 107072 (2020).
[7] F. Zonca, L. Chen, S. Briguglio, et al., New Journal of Physics 17, 013052 (2015).
[8] F. Zonca, F. Romanelli, G. Vlad, and C. Kar, Phys. Rev. Lett. 74, 698 (1995).
[9] L. Chen, F. Zonca, R. Santoro, and G. Hu, Plasma Physics and Controlled Fusion 40, 1823 (1998).

Speaker: Ningfei Chen (Max-Planck Institute for Plasma Physics)

25 Optimization of active control of ICRH-driven TAE using 3D magnetic perturbations

In magnetically confined fusion devices, energetic particles produced by external heating systems such as neutral beam injectors (NBI) or ion cyclotron resonance heating (ICRH) may excite MHD instabilities such as Alfven Eigenmodes (AEs) which can degrade confinement in tokamak plasmas and potentially damage the plasma facing components [1]. Externally applied 3D Magnetic Perturbations (MPs) have previously been proposed as an effective actuator for controlling NBI-driven AEs in several devices, including ASDEX Upgrade [1,2], KSTAR [3], EAST [4,5] and NSTX [6,7].

During the 2024-2025 and 2025-2026 ASDEX Upgrade experimental campaigns, a series of discharges with B0 = -2.5 T and Ip=500 kA incorporating minority-H ICRH heating and counter current ECCD were to investigate the effect of the n=2 and n=4 MPs on Toroidal Alfven Eigenmode (TAE) stability. The n=2 minority-H ICRH-driven TAE amplitude has been tracked along a varying differential phase for the MP and TAE mitigation was observed for an optimal value of the differential phase near 0°. The results also show that the modulation is independent of the absolute phasing of the MP coils for ICRH-driven TAE as opposed to NBI-driven TAE.

In this work, we present these experimental results supported by numerical modelling with MEGA [8] and ASCOT [9]. The simulations predict the appearance of unstable ICRH-driven TAEs across various possible values of energetic particle anisotropy and pressure. The TAE modulation found in experiments has been numerically investigated with a set of simulations incorporating MPs of differing differential phases and toroidal numbers. These simulations also incorporate ICRH energetic particles, which are modelled as a highly anisotropic hydrogen slowing down distribution function. This aims to provide a drive for the TAE while also being able to resonate with the MP fields.

[1] M.Garcia-Munoz et al., Plasma Phys. Control. Fusion 61 054007 (2019)
[2] J.Gonzalez-Martin et al., Phys. Rev. Lett. 130, 035101 (2023)
[3] K.Kim et al., Nucl. Fusion 60 126012 (2020)
[4] N.Chu et al., Nucl. Fusion 58 104004 (2018)
[5] J.Zhao et al., Plasma Sci. Technol. 095101 (2021)
[6] G.J.Kramer et al., Plasma Phys. Control. Fusion 58 085003 (2016)
[7] A.Bortolon et al., Phys. Rev. Lett. 110 265008 (2013)
[8] Y.Todo et al., Phys. Plasmas 5, 1321–1327 (1998)
[9] K. Särkimäki et al 2016 Plasma Phys. Control. Fusion 58 125017

This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200 — EUROfusion). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.

Speaker: Daniil Kabirov (University of Seville)

26 Perturbed bootstrap current of alpha particles and its contribution to tearing-mode stability

The interaction between fusion-born alpha particles and resistive MHD modes is a key stability concern for burning plasmas such as ITER. The neoclassical tearing mode — one of the most performance-limiting instabilities in these regimes — is itself driven by the helical perturbation of the bootstrap current associated with pressure flattening inside magnetic islands, which makes the perturbed bootstrap current the central quantity governing its stability. An additional perturbed bootstrap current carried by alpha particles can therefore directly affect tearing-mode and neoclassical-tearing-mode (NTM) stability. In our earlier work, the influence of the perturbed alpha-particle pressure was treated through its modification of the Mercier-index term; the perturbed bootstrap current carried by alpha particles constitutes a distinct and so-far untreated channel, and is the subject of the present study.

We carry out a drift-kinetic derivation, within the framework of Graves (2013), of the non-flux-function perturbed parallel current produced by an isotropic alpha-particle population, treating trapped and passing particles on an equal footing. Starting from the microscopic perturbed distribution function $\delta g$, the two classes are found to behave qualitatively differently. For trapped particles, the leading-order contribution cancels upon summation over the sign of the parallel velocity, and the surviving contribution originates from the first-order finite-orbit-width correction; it carries the structure of a perturbed bootstrap current. For passing particles, the resonance denominator depends on the parallel-velocity sign, so a non-zero contribution already appears at leading order. The two channels enter the tearing-mode outer-region equation with different singularity structures near the rational surface.

These results provide the analytic source terms for an ongoing quantitative evaluation of the alpha-particle contribution to $\Delta'$. The same drift-kinetic framework applies equally to NBI- and ICRF-driven energetic ions, and the results bear directly on the assessment of energetic-particle effects on neoclassical-tearing-mode onset and control in future burning-plasma devices.

Speaker: xiaoxi zhang

27 Reduced model of the interaction between microturbulence, energetic particles and energetic particle driven modes.

We present a reduced framework to describe the multiscale interaction between thermal plasma microturbulence, energetic particles generated by auxiliary heating, and the macroscopic instabilities that these particles can drive. Our approach focuses on how microturbulence regulates the saturation amplitude of energetic-particle-driven modes, such as shear Alfvén waves, through scattering-dominated nonlinear dynamics, where effective particle scattering balances wave dissipation. At the same time, energetic-particle-driven modes can modify the underlying microturbulence via nonlinear couplings and zonal mode excitation, introducing a feedback loop that can reduce turbulence levels and, in turn, lower the mode saturation amplitude. This interplay may lead to improved thermal plasma confinement and reduced energetic particle transport. The proposed framework provides a tractable way to capture the essential physics of this coupled system and to explore the mutual interaction between turbulence, energetic particles, and macroscopic instabilities.

Speaker: Tommaso Barberis (Princeton Plasma Physics Laboratory)

28 Self-consistent Interactions Between Full-energy Helium Ions and a Drift Wave-Zonal Flow System in Burning Plasmas

Future D–T burning plasmas generate energetic alpha particles that gradually evolve into thermalized helium ash. The resulting helium ions modify turbulence through dilution effects, while turbulence simultaneously regulates helium-ion transport. Understanding this mutual interaction is essential for predicting both confinement performance and helium-ash exhaust in future fusion reactors.

A self-consistent one-dimensional spatiotemporal model is developed to investigate the interactions between full-energy helium ions and a drift wave-zonal flow (DW–ZF) system. The model self-consistently evolves helium-ion dilution, drift-wave turbulence and zonal flows, while incorporating dilution effects on both the drift-wave frequency and growth rate. Unlike conventional approaches employing prescribed dilution profiles, the present framework captures the nonlinear feedback between helium ions and turbulence.

Numerical results show that the self-consistent evolution significantly reduces the saturated helium-ion dilution factor, particularly the contribution from lower-energy helium ash. Consequently, helium-ash removal becomes more achievable, while drift-wave turbulence is enhanced and zonal-flow energy is reduced. These findings suggest that the beneficial role of dilution in improving confinement may be overestimated in models employing fixed dilution profiles. Overall, this work highlights the importance of self-consistent helium-ion/turbulence evolution for reliably assessing burning-plasma performance.

Reference
Z. Mai, W. Guo* and L. Wang, Nuclear Fusion 66 (2026) 026004

Speaker: Weixin Guo (Huazhong University of Science and Technology)

29 The saturation mechanism of fishbone

Energetic particles (EPs) are essential for sustaining burning plasmas through self-heating, but they can also drive global instabilities that degrade EP confinement, making quantitative prediction of EP transport crucial for achieving high fusion performance. Among these instabilities, fishbone modes are particularly important because their macroscopic mode structures can induce significant EP transport. Since the level of EP transport is largely determined by the nonlinear saturation amplitude of the fishbone mode, understanding its saturation mechanism is a key issue. The conventional saturation mechanism is associated with resonant wave-particle interaction, which flattens the EP distribution gradient in phase space. However, recent gyrokinetic simulations have suggested that fishbone-induced zonal flows may play a dominant role in the nonlinear saturation of fishbone modes.
In this work, nonlinear global gyrokinetic simulations are performed using the optimized GTC code to simulate fishbone dynamics, including a more self-consistent treatment of zonal fields. In particular, by improving the zonal-field equations to include the contribution from the zonal electron density perturbation, we find that the saturated zonal-flow amplitude is significantly reduced compared with simulations in which this contribution is neglected. As a result, the self-generated zonal flow does not appear to be the dominant saturation mechanism of the fishbone mode in our simulations. When the zonal fields, including zonal flow and zonal current components, are artificially suppressed, the fishbone mode can still reach nonlinear saturation with a saturation amplitude comparable to that obtained in the fully self-consistent simulation. These results indicate that the dominant fishbone saturation mechanism is more likely associated with nonlinear EP dynamics, such as resonant phase-space redistribution, while fishbone-induced zonal fields play a subdominant role under the present conditions.

Speaker: Huisan Cai (University of Science and Technology of China (CN))

30 Simulation results of energetic particle driven instability in a tokamak configuration

Energetic particles (EPs) include fusion alpha particles and the fast ions from auxiliary heating. These particles can drive instabilities in a toroidal plasma. The instabilities then push the EPs outward. This lowers the heating efficiency and raises the heat load on the first wall.

These instabilities are studied with the MEGA code. MEGA is a hybrid code: the bulk plasma is a nonlinear MHD fluid, the EPs are treated kinetically, and the two are coupled through the EP contribution to the MHD momentum equation. The equilibrium and the plasma profiles come from ASDEX Upgrade (AUG) discharge #36267 at t = 4.09 s. The EP distribution is an anisotropic slowing-down distribution, Gaussian in pitch angle. The NBI energy is 93 keV, and the EPs are born at about half the Alfven speed.

One dominant instability is found in the plasma core. Its mode numbers are m/n = 3/1 and its frequency is 0.11 times the Alfven frequency, about 60 kHz. The mode appears only above a threshold in the EP beta, between 0.2% and 0.4%. Between 0.4% and 1.2% the growth rate rises steadily, while the mode structure and the frequency hardly change. Above 1.5% the growth rate becomes larger than the frequency, and the frequency itself falls by 17%. The mode is then no longer a weakly driven eigenmode. The numerical convergence was also checked. With eight times more markers, the growth rate changes by 0.4% and the frequency by 1.1%.

None of these scans reproduced the frequency measured in the experiment. The reason lies in the equilibrium. The safety factor q in the plasma core is the least reliable part of the reconstruction. The same equilibrium gives a core q of 2.05 or 2.40, depending on how it is computed. The reconstructed profile has a small maximum of q close to the magnetic axis. This maximum was removed, and the simulations were repeated. The mode is still there, and its growth rate is almost the same. Its frequency, however, moves up by 52%. The core q profile therefore controls the frequency of the mode, but not its existence. This correction is what brought the simulations close to the experiment.

Two of the simulated frequencies then agree with the measurement. The radial position of the EP pressure peak was scanned on both equilibria. The two scans give almost the same results, except at one position. When the EP pressure peaks near mid-radius, only the modified equilibrium gives a mode at 0.0817 times the Alfven frequency, which is 44.89 kHz. The original equilibrium gives a different mode there. This match therefore appears only after the core q maximum is removed. When the EP pressure peaks further out, both equilibria give the same mode at 0.1896 times the Alfven frequency, which is 104.14 kHz. This mode is carried mainly by the m = 4 harmonic. It lies in the m = 3/4 continuum gap, but it also crosses the continuum, so it is not a gap mode in the strict sense. These two frequencies, 44.89 and 104.14 kHz, are the strongest link between the simulations and the experiment so far.

Next, the nonlinear behaviour and the redistribution of resonant particles will be examined, and a detailed comparison with the experiment will be made.

Speaker: Dr Hao Wang (National Institute for Fusion Science)

EP impact on machine operations and safety

31 Tungsten sputtering due to fast particles in ITER

Fast-ion interactions with plasma-facing components remain an important open question for ITER, particularly after the recent transition to a full tungsten first wall. Tungsten sputtering is expected to be mostly caused by the thermal plasma, but lost fast ions such as fusion-born alpha particles may contribute to tungsten sputtering. The extent of fast ion contribution depends on both the spatial distribution of fast-ion losses and the corresponding sputtering yields, which vary with particle energy and angle of incidence. Previous studies indicate that sputtering yields decrease at higher energies, while for high-energy ions the yield increases with incidence angle. These considerations raise the central question of whether fast-ion losses can produce a significant tungsten source compared to thermal processes.

We combine ASCOT-based fast-ion wall interaction data with sputtering yield calculations derived from SDTrimSP for hydrogenic species and helium (alphas). High-resolution yield maps as functions of energy and incidence angle were constructed to enable numerical integration with fast-ion flux distributions. The methodology converts alpha particle fluxes to wall source rates and subsequently to tungsten sputtering rates by applying energy- and angle-dependent yields. These results are then compared against thermal sputtering contributions obtained from WallDyn modelling. Furthermore, we evaluate numerically the impact of CX lost alphas and NBI on tungsten sputtering under realistic ITER conditions.

The results indicate that the tungsten source due to fast alpha particles is significantly smaller than that from thermal processes, with differences on the order of 3–4 orders of magnitude. This large gap is attributed to both relatively low sputtering yields under relevant conditions and limited fast-ion fluxes to the wall. No clear mechanism was identified that could bridge this discrepancy under ITER-relevant conditions, and contributions from beam ions are expected to be even smaller.

Speaker: Antti Snicker (VTT Technical Research Centre of Finland Ltd.)

32 Quantitative assessment of ICRF antenna limiter tiles performance under sustained fast ion heat loads on ASDEX Upgrade

This study is motivated by the observation of a macroscopic (~10 cm in scale) carbon tile failure in the ion cyclotron range of frequencies (ICRF) limiter structure during a hydrogen minority ion cyclotron heating scenario in D plasma on ASDEX Upgrade (AUG). In this discharge (#42762): the on-axis magnetic field is 2.65 T, the plasma current is 0.36 MA, the core plasma electron density is 5-6x10^19 m^-3, the hydrogen minority fraction is ~5%, and the ICRF operational frequency is 36.5 MHz. As the ICRF power increased from 0.15 MW to 3.3 MW between 1.30 s and 7.95 s, a part of the ICRF antenna protection limiter tile in Sector 12 (near the midplane) was seen to separate just before 8 s. The tile failure necessitated an unscheduled torus opening to replace the broken tile.
In order to estimate heat loads induced by lost energetic ions on the ICRF limiter structure, we employ ASCOT-RFOF. This code has earlier demonstrated its applicability in simulating ICRF heating and fast ion loss detector signals on AUG [1]. The initial ASCOT-RFOF simulation runs revealed a spatially concentrated heat load pattern, where the values range from several MW/m^2 to nearly zero across a distance of just a few cm. To resolve such narrow heat loads, the entire AUG outer wall surface mesh required an upgrade from the previous version with mesh triangles measuring ~10 cm across to a more refined mesh with triangles ~1 cm across. CAD data was used to generate a new ASCOT 3D wall mesh for all plasma-facing components of interest, in particular the four ICRF antennas with their limiters and Faraday screens. Applying the new high-resolution mesh and a high fast ion marker count (~10 million) to the studied plasma scenario reveals heat load values that reach 10 MW/m^2 (at highest ICRF power) in a highly concentrated tile surface region, just a few cm across. The tile regions experiencing these high heat loads are on the ‘right hand’ side of the antenna limiter (when viewed from the plasma), in the vicinity of the Faraday screen rods. The location is well matched to the sheared-off section of the midplane tile #6. The ASCOT output reveals additional hot spots located on the Faraday screen, near the tile surfaces, and on the ‘left hand’ side of the limiter. The simulated heat loads are currently being coupled to ANSYS, a thermo-mechanical analysis software, to estimate material stresses on the antenna limiter tiles. The overall computational loop that links plasma-generated fast ions with thermo-mechanical stresses experienced by the outer wall structures will be presented, in view of applying such a computational workflow to future fusion devices such as ITER.
For details of the ASCOT wall heat load simulation model, see the contribution of S. Sipilä et al. in this Technical Meeting.
[1] S. Sipilä et al., 2021 Nucl. Fusion 61 086026.

Speaker: Roman Ochoukov (IPP - Max-Planck-Institut für Plasmaphysik (EU))

33 Fast-ion heat loads due to static 3D fields and diagnostic design for tokamak power plant

Future tokamak power-plant prototypes will rely on alpha heating to sustain burning-plasma conditions. However, the transport and loss of a minority of energetic alpha particles can lead to significant heat loads on plasma-facing components (PFCs), which must withstand unprecedented steady-state particle and power fluxes during operation.
The development and validation of reliable workflows to predict alpha-particle heat loads and assess the viability of associated diagnostics are therefore essential. To quantify these effects, the LOCUST[1] (Lorentz Orbit Code for Use in Stellarators and Tokamaks) full-orbit code is employed to model steady-state heat loads on PFCs. LOCUST tracks fast-ion trajectories in the presence of three-dimensional (3D) magnetic perturbations and includes Monte Carlo collisions with the background plasma.
The workflow is compared against fast-ion loss detector (FILD) measurements of alpha particles from deuterium–tritium (DT) discharges in JET [2]. The contributions from both beam–target and thermonuclear reactions are modelled during the discharge. While the steady-state heat flux on the poloidal limiters and other structures is found to be low and difficult to distinguish from other sources in infra-red measurements. Synthetic diagnostic signals generated for the FILD show strong agreement with experimental observations.
The workflow is subsequently applied to STEP[3] (Spherical Tokamak for Energy Production), where the impact of error fields arising from toroidal-field (TF) coil misalignments is analysed alongside magnetic perturbations generated by edge-localised-mode control coils (ELMcc). The results show that, for compact outboard configurations with an external coil radius of 9 m, relatively small TF-coil misalignments of only a few centimetres can produce excessive alpha-particle power fluxes exceeding 5 MWm⁻². In contrast, larger misalignments can be tolerated when the coil radius is increased to at least 10.5 m, for which peak power fluxes remain below 0.2 MWm⁻². While different ELMcc current configurations result in varying levels of alpha-particle transport, the total loss of alpha-heating power remains low, with peak heat fluxes reaching up to 0.9 MWm⁻² on the lower divertor and 0.32 MWm⁻² on the first wall. These results also demonstrate the strong influence of the plasma response to externally applied magnetic perturbations. Furthermore, the presence of high-density, low-temperature plasma in the scrape-off layer is found to provide a significant protective effect to the first wall. Finally, preliminary simulations exploring the implementation of a FILD diagnostic for monitoring alpha-particle losses in STEP are presented.

Speaker: Fabio Camilo de Souza (UKAEA)

18:30 Reception

Dining reception for the EP-TM 2026 participants

Tuesday 29 September

EP in stellarators and 3D configurations

34 Energetic particle instability characteristics of reactor-scale stellarators

Recently stellarators have made significant progress with respect to improvements in the classical guiding center energetic particle confinement via shape optimization [1]. However, as has been observed with tokamaks once the neoclassical confinement is reduced to a low level, the confinement properties of energetic particle (EP) populations may be more dominated by instability-induced transport phenomena [2]. In particular, the various Alfvén eigenmodes (AE) that can be resonantly destabilized by EP components provide channels for enhanced transport of EPs. To address this issue, the FAR3d gyro-Landau closure model [3] has been employed. This is a reduced model that accurately models the mode coupling physics which is important to the analysis of these instabilities in 3D configurations, while using a simplified parallel EP resonance closure model to excite the instabilities. This provides a means to rapidly evaluate the linear stability of many configurations and scenarios and a feasible method to study the nonlinear saturation and resulting transport, considering interactions between multiple modes. Stellarators offer the possibility to operate at high density regimes where the slowing-down time is decreased, resulting in lower drives for EP instabilities. This effect is analyzed for a reactor-sized quasi-helical stellarator, Wistell-D, through evaluating linear stability along constant fusion power contours (0.5, 1, 2, and 3 GW) in electron density and temperature space for the n = 1, 2, and 3 mode families. For this example, suppression of AE’s is seen up to 1 GW, but residual modes persist at higher power levels, even at high densities. This work is then extended into the nonlinear regime via simulations of the Wistell-D and a reactor scaled version of W7-X; toroidal modes up to n = 30, are followed to saturation. Zonal flow/current generation and mild flattening of the EP density profile are observed from which global EP particle and energy fluxes are derived. The flux-gradient relations exhibit non-local characteristics, as has been seen in applications of FAR3d to tokamaks [4], and motivates going beyond the usual diffusive and critical gradient models. This type of EP transport can be characterized using machine-learning surrogate models which can interface with integrated simulation models. The efficient performance of FAR3d and its demonstrated application to a wide range of stellarator configurations also motivate its use as a target function for stellarator shape optimization to further explore possibilities for suppression of AE instabilities in 3D configurations.

[1] M. Landreman, S. Buller, and M. Drevlak. “Optimization of quasi-symmetric stellarators
with self-consistent bootstrap current and energetic particle confinement,” Physics of Plasmas 29.8
(2022).
[2] W. W. Heidbrink, N. N. Gorelenkov, Y. Luo, M. A. Van Zeeland, “Anomalous Flattening of the Fast-Ion Profile during Alfvén-Eigenmode Activity,” Phys. Rev. Lett., Vol. 99, 245002 (2007).
[3] J. Varela, D. Spong, L. Garcia, Y. Ghai, J. Ortiz and FAR3d project collaborators, “Stability optimization of energetic particle driven modes in nuclear fusion devices: the FAR3d gyro-fluid code,” Frontiers in Physics, 12:1422411 (2024)
[4] D.A. Spong, Y. Ghai, J. Varela, L. Garcia, "Nonlinear Alfven instability simulation and EP transport for ITER reversed shear (steady-state) and monotonic q-profile regimes," Nuclear Fusion, 2025; 61(11):116061.

Acknowledgements - This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, under Award DE-AC05-00OR22725.

Speaker: Don Spong (Oak Ridge National Laboratory)

35 Effects of Pressure Gradients of Multiple Fast-Ion Species on Fast-Ion-Driven Instabilities in the Large Helical Device

In magnetic confinement fusion devices, there are multiple fast-ion species such as fusion product (i.e., α particles) and fast deuterons produced by neutral beam injection. These fast ions may induce MHD instabilities such as Alfvén Eigenmodes (AEs) which can enhance fast-ion transport and losses. The study of fast-ion-driven MHD instabilities is one of the important topics for fusion reactors.
In order to clarify the relationship between instabilities and fast ion transport in the plasma with multiple fast ion species, experiments with the combined injection of hydrogen and deuterium beams were conducted in the Large Helical Device (LHD). In these experiments, recurrent AE bursts caused by the fast protons and fast deuterons were observed [1].
Computer simulation is an effective approach for investigating instabilities driven by multiple fast-ion species. A hybrid simulation code for nonlinear MHD and energetic-particle dynamics, MEGA, has been developed to simulate recurrent bursts of fast-ion-driven MHD instabilities, including sources, collisions, and losses [2]. In MEGA simulations for LHD experiments with fast protons and fast deuterons, AEs with frequencies consistent with experimental observations were successfully reproduced. In addition, it was shown that the synergistic effects of multiple fast-ion species enhance fast-ion transport and losses in LHD [3].
On the other hand, in the previous simulation conditions, AEs could not be destabilized only by fast deuterons. The synergistic effects of multiple fast-ion species on AEs and fast-ion transport remain unclear when the fast-ion pressure gradients vary, particularly in cases where the fast-deuteron pressure gradient becomes sufficiently large to excite AEs.
In this study, MEGA simulations of AEs driven by multiple fast-ion species are performed in the LHD magnetic field configuration when the pressure gradients of fast deuterons and fast protons are varied. In the LHD, as the fast deuteron pressure gradient increased, the amplitude of the initial AEs became larger. In addition, the amplitudes of the additional AEs associated with the redistribution of fast protons and fast deuterons also increased, accompanied by enhanced fast-ion transport. The energy transfer between the AEs and fast ions will be evaluated and the dependence of AEs driven by multiple fast-ion and its transport on the fast-ion pressure gradients will be shown.

[1] S. Kamio, et al, Plasma and Fusion Research, 16, 2402044 (2021).
[2] Y. Todo, et al., Nucl. Fusion 55, 073020 (2015).
[3] R. Seki, et al., Proceedings of 30th IAEA Fusion Energy Conference, Chengdu (2025).

Speaker: Ryosuke SEKI (National Institute for Fusion Science)

36 Properties of Magnetohydrodynamics Instabilities in ITER-Scale Helical Core Plasmas

The tokamak hybrid scenario is a candidate for long-pulse, sawtooth-free operation in ITER. This sawtooth-free state is attributed to its robust, near-flat q≳1 core safety factor profile, maintained by the long-lived MHD-driven current redistribution that hinders the peak current profile. A notable example is the long-lived m/n=1/1 quasi-interchange mode. However, this m/n=1/1 mode leads to a long-lived core toroidal asymmetric state, also known as “helical core” (HC)1. Studies predicted that a large HC could spontaneously form in ITER2. Its non-negligible n=1 toroidal asymmetry may affect the overall plasma stability, as all toroidal modes are linearly coupled3-4 into a single entity. In this study, we assess the MHD stability of ITER-scale HC equilibrium using the kinetic MHD energetic-particle (EP) simulation code. We find that this toroidal asymmetry leads to the destabilization of the moderate-to-short-wavelength pressure-driven and EP-driven MHD modes, which otherwise are stable in the axisymmetric equilibrium. Both modes consist of broad toroidal and poloidal spectra. All toroidal harmonics of the mode balloon in the bad curvature region, but they are aligned such that they constructively interfere along the HC compressed flux region and destructively interfere elsewhere. For the pressure-driven mode, its linear growth rate follows the resistive ballooning mode scaling, which may suggest that it is a cluster of “n” ballooning modes synchronized by HC. For the EP-driven mode, it resides in the Alfvenic-acoustic gap. Unperturbed orbit analysis shows that the toroidal sideband resonances of this mode overlap spatially, which may explain its destabilization in HC equilibrium. The nonlinear simulation predicts that these pressure-driven and EP-driven MHD modes lead to a minor redistribution of the bulk plasma and fusion-born alpha particles. Nonetheless, this quantitative prediction should be interpreted cautiously, as the scale length of the short-wavelength spectra of the modes may be comparable to or smaller than the ion gyroradius, and they are also sensitive to viscous–resistive dissipation imposed in our model.
[1] W.A. Cooper et al 2010 Phys. Review Letters 105.3 035003
[2] A. Wingen et al 2018 Nucl. Fusion 58 036004
[3] Y.I. Kolesnichenko et al 2002 Phys. of Plasmas 9.2 517-528
[4] D. Spong et al 2003 Phys. of Plasmas 10.8: 3217-3224

Speaker: Dr Panith Adulsiriswad (adulsiriswad.panith@qst.go.jp)

37 Resonant plateau collisional passing alpha losses

Alpha particles typically exhibit collisionless dynamics throughout most of their phase space. However, their motion can be very sensitive to collisions under resonant conditions. The collision operator is required to resolve the singular behavior around resonances, allowing the formation of a collisional boundary layer, where a perturbed distribution function forms and drives radial transport. In an optimized quasisymmetric (QS) stellarator, the presence of an error field of mode numbers m,n results in a passing alpha particle resonance near the q=m/n rational surface, for those alphas whose streaming motion and tangential drift cancel. The resonance can extend over much of the radial cross section because the resonant velocity space pitch angle varies with minor radius. If the error field is sufficiently small, the collisional boundary layer around the resonance is wider than the island structure due to the drift and streaming. For larger error fields, the drift and streaming island structure prevails and the collisional boundary layer localizes around the island separatrix. In this work, we develop a drift kinetic model for alpha particles that enables an evaluation of passing alpha particle and energy collisional transport. We quantify the associated diffusivities and resulting alpha particle and energy losses, and show that these can be significant. The calculation is performed using a general QS stellarator background with an error field that causes a deviation from QS. The results are also applicable to a tokamak. In the smaller error field limit, our model predicts energy losses of up to 10% for error fields of order \delta B=10^(-3) T in a background magnetic field of 10 T.

Speaker: Miguel Calvo Carrera (Massachusetts Institute of Technology)

10:50 Coffee break

Physics of EP modes and transport

38 Alfven eigenmodes driven by α particles in JET D-T plasmas

The Joint European Torus (JET) recently completed historic deuterium-tritium (D-T) campaigns, which provided unique opportunities to study burning plasma physics. Alfvén eigenmodes were excited by fusion-born α particles [1] in several high-performance plasmas with fusion power > 8 MW externally heated only by neutral beam injection at up to 32 MW [2]. These modes were absent in pure deuterium reference plasmas, indicating that α particles were the source of instability. Magnetic probe signals were sufficiently strong to determine the toroidal mode numbers n, revealing that both axisymmetric ($n=0$) and non-axisymmetric ($n=1$) harmonics were destabilised. The excitation of the axisymmetric mode indicates drive from positive energy gradients in the α particle distribution, distinct from the expected drive from radial gradients.

Modelling reveals that the electric field of both modes peaked at the high-field-side edge of the plasma. The modes were driven by counter-current passing α particles with large orbit widths that enabled α particles born in the core to interact with modes at the plasma edge. The drift of the counter-passing α particles led to significant spatial overlap between the high-field-side orbit excursion of the α particle and the electric field of the eigenmodes, enabling the efficient power transfer that destabilised the Alfvén eigenmodes. Conversely, damping from beams ions was relatively weak because the co-passing beam ions had narrow orbit widths and drifted toward the low-field-side of the tokamak, away from the eigenmode. The high magnetic field in this plasma (3.85 T on the magnetic axis) and mode location on the high-field-side increased the Alfvén velocity and limited Landau damping by the thermal and beam ions to inefficient sideband resonances.

The drive from α particles came from positive energy gradients at the edge of the plasma – a universal feature of all magnetically confined plasmas due to the minimum energy required to confine particles. To explore the implications for future burning plasmas, this presentation will compare the α particle drive of the axisymmetric mode, which was entirely from energy gradients, with the drive of the non-axisymmetric mode, which accessed energy from gradients in the toroidal angular momentum $P_\phi$ in addition to energy gradients.

References
[1] H. J. C. Oliver et al., PRL, 136, 055101 (2026)
[2] D. B. King et al., PPCF, 68, 045023 (2026)

Speaker: James Oliver (UKAEA)

39 Nonlinear evolution of perturbatively driven far-from-threshold kinetic instabilities

The kinetic destabilization of gap modes, e.g., toroidal Alfvén eigenmodes, results in both diffusive and convective transport of energetic particles [1]. In scenarios where fast relaxation events result in very rapid redistribution of fast ions or rapid changes to the mode resonance conditions, a mode can be excited in a regime far from the instability threshold. For example, strong Alfvén eigenmode activity has been observed in both JET [2] and JT-60U [3] directly following sawtooth crashes, which can redistribute fast particles on a timescale faster than the mode growth rate. While the wave saturation level of these interactions has been established in the strongly driven regime within a perturbative approximation [4,5], no analytical description of the time evolution of the instability has been developed. We examine the dynamics of a strongly driven kinetic instability at a single resonance in the presence of collisional and dissipative processes. The wave evolution can be approximated as occurring in two phases: first, linear growth driven by the positive distribution gradient, then slower, weakly nonlinear evolution driven by the balance of the wave drive and dissipation. In the second phase of evolution, we find that the distribution evolves time locally with the evolution of the mode, and by exploiting this time locality, we construct a piecewise-continuous, closed-form analytical solution for the time evolution of the mode amplitude. This result agrees closely with nonlinear kinetic simulations performed using the BOT code [6].

[1] W. W. Heidbrink, Phys. Plasmas 15, 055501 (2008).
[2] J. Ruiz Ruiz et al., Phys. Rev. Lett. 134, 095103 (2025).
[3] G. J. Kramer et al., Nucl. Fusion 41, 1135 (2001).
[4] H. L. Berk and B. N. Breizman, Phys. Fluids B 2, 2226 (1990).
[5] N. Petviashvili, Ph.D. Thesis, University of Texas at Austin (1999).
[6] M. K. Lilley et al., Phys. Plasmas 17, 092305 (2010).

Speaker: Emma Devin (Princeton University)

40 Impact of toroidal Alfvén eigenmodes and sawtooth crashes on NBI fast ions on ST40

ST40 is a high-field spherical tokamak with a major radius up to 0.55 m and toroidal magnetic field up to 2.1 T, operating at plasma currents ranging from 0.2-1.0 MA. It is equipped with two neutral beam injectors (NBI) with 1.0 MW at 55 keV and 0.8 MW at 24 keV. Across the wide toroidal field range the fast ions from NBI can range from sub- to super-Alfvénic, driving a variety of Alfvénic modes including toroidal Alfvén eigenmodes (TAE). These are commonly observed at the start of the discharges but are often suppressed later in the flat-top, in particular after sawtooth oscillation begins.

In this contribution we present experimental and modelling results on interaction between NBI fast ions and TAEs and sawteeth. Discharges have been analysed across the ST40 operating range where TAEs have been observed. The fast particle code ASCOT has been used to assess the drive from fast ions and fast ion transport in various scenarios. While prevalent, TAEs on ST40 typically do not appear to cause significant fast ion losses based on measured neutron rates and neutral particle analyser (NPA) fluxes.

In contrast, sawtooth crashes are often associated with dips in neutron and NPA measurements. The effect of sawteeth has been investigated with ASCOT using an ad-hoc model of flattening the fast ion distribution within the mixing radius. These results suggest significant fast ion redistribution, potentially suppressing the drive for TAEs. This is also supported by integrated modelling with ASTRA, where anomalous fast ion diffusion up to 1 m²/s is required to match the experiment also in the absence of TAEs.

Speaker: Dr Jari Varje (Tokamak Energy Ltd)

12:45 Lunch break

Poster session II and Coffee break

41 The resilience of Elliptical Alfvén Eigenmodes (EAEs) in high-pressure tokamaks

Transport of alpha-particles born in deuterium-tritium (D-T) nuclear fusion reactions is a key issue for magnetic fusion. Wave-particle resonant interactions – between Alfvén waves and alpha-particles born at super-Alfvénic energy – could induce significant radial re-distribution of alpha-particles, affecting both the self-heating efficiency and integrity of the first wall [1].

The effect of high thermal plasma pressure gradients on the existence of toroidal Alfvén eigenmodes (TAEs), which was first considered theoretically in [2], could potentially present a very attractive option for resolving the concerns raised by the presence of TAEs in burning fusion plasmas. This effect is not associated with an increase in TAE damping caused by kinetic interactions, but is due rather to the fact that ideal MHD TAEs cease to exist above a certain critical plasma pressure gradient. Here we investigate the thermal plasma pressure effect on Elliptical Alfvén Eigenmodes (EAEs) – which could be particularly important for the spherical tokamak route to fusion since these devices have high natural ellipticity e and high normalized pressure.

Numerical investigations performed using the MISHKA MHD eigenvalue code on JET and ST40-like discharges reveal a higher critical pressure gradient for the non-existence of EAEs than that of TAEs. Building on the analysis of [3,4], a new analytic theory applicable to core-localised EAEs and developed in the limit e/S >>1 (where S is magnetic shear) will be presented.

[1] ITER Physics Expert Group on Energetic Particles, Heating and Current Drive and ITER Physics Basis Editors Nucl. Fusion 39, 2471 (1999)
[2] Fu G Y and Cheng C Z 1990 Physics of Fluids B: Plasma Physics 2 985–993
[3] Berk H L et al 1995 Physics of Plasmas 2 3401
[4] Candy et al Physics Letters A 215 (1996) 299

This work has been funded by the EPSRC Fusion Grant 2022/27 [grant number EP/W006839/1]. We thank Michael Fitzgerald, Daniele Brunetti and Ken McClements for several helpful comments.

Speaker: Arkaprava Bokshi (UKAEA)

42 Gyrokinetic Simulations for Spherical Tokamak Energetic Particle-Turbulence Experiments

Gyrokinetic Simulations for Spherical Tokamak Energetic Particle-Turbulence Experiments

C. Gallaro¹†, T. Barberis¹, T. Adkins¹, P. J. Bonofiglo¹, and V. N. Duarte¹

¹Princeton Plasma Physics Laboratory, Princeton, New Jersey, USA
†cgallaro@pppl.gov

Recent experiments on the MAST-U spherical tokamak investigated the multi-scale coupling between energetic particles, Alfvénic eigenmodes (AEs), and microturbulence [1]. In particular, the experiments consisted of using neutral beam injection to create sequences within the same discharge where there is a clear presence and absence of beam ions to probe the two regimes. This multi-scale relationship has been examined in multiple advanced tokamaks such as JET [2] and DIII-D [3], demonstrating the pumping of a zonal mode via beating of toroidal Alfvén eigenmodes leading to an increase in thermal confinement. However, a precise coupling has not been quantified in a spherical tokamak where low aspect ratio and high-𝛽 alter both the AE spectrum and the nature of the dominant microinstabilities. Here we present gyrokinetic simulations using the CGYRO [4] code to determine the spectrum of turbulent fluctuations of both electrostatic and electromagnetic character for the MAST-U plasmas of interest. Based on previous studies, we expect that the dominant contributions from the linear spectrum will be microtearing modes [5], depending on the equilibrium flow shear and collisionality, as well as high-k electron-scale turbulence [6]. These simulations will be compared with experimental data collected via the Doppler backscattering system to verify the nature of the fluctuations present in the discharges. We will then compare these spectra in the presence and absence of beam ions to determine the primary direct and indirect effects of the beam ions on the turbulence. Lastly, we consider the inclusion of energetic particles in linear CGYRO simulations to comment on the spectrum of Alfvénic fluctuations excited by beam ions in the aforementioned MAST-U plasmas.

References
[1] P. Bonofiglo et al., 2026, in 19th Technical Meeting on Energetic Particles in Magnetic Confinement Systems
[2] J. Ruiz Ruiz et al., 2025 Phys. Rev. Lett. 134 095103
[3] X. D. Du et al., 2025 Phys. Rev. Lett. 135 265101
[4] J. Candy, E. A. Belli, and R. V. Bravenec, 2016 J. Comput. Phys. 324 73
[5] M. Giacomin et al., 2023 Plasma Phys. Control. Fusion 65 095019
[6] D. C. Speirs et al., 2025 Nucl. Fusion 65 046019

Acknowledgements
This work is based upon work supported by the US Department of Energy, Office of Science, Office of Fusion Energy Sciences, and has been authored by Princeton University under Contract DE-AC02-09CH11466 with the US Department of Energy. The work was supported by the DOE Early Career Research Program, project "Phase-Space Engineering of Supra-Thermal Particle Distribution for Optimizing Burning Plasma Scenarios".

Speaker: Cosmo Gallaro

43 Measurement and validation of sawtooth induced fast ion transport on DIII-D

Sawtooth oscillations and the associated fast ion transport play a critical role in plasma stability, heating efficiency, and fusion performance. Despite general similarities, sawteeth show a broad variance in characteristics, such as sawtooth crash time, amplitude, and period, depending on the safety factor q, plasma current, plasma beta, plasma shape, and collisionality. The improved diagnostics, especially Imaging Neutral Particle Analyzers (INPAs) and Imaging Fast-Ion D-Alpha (IFIDA) fast-ion diagnostics, enable a more refined study of sawteeth dynamics and their impact on fast-ion transport in phase space. An empirical sawtooth database has been built to characterize sawtooth behavior and quantify the resulting fast-ion transport. The database confirms the general observation that passing fast ions are often redistributed by sawteeth from the plasma core to regions outside the q=1 surface, leading to reductions in central fast-ion density of up to 30% in the core. In contrast, trapped fast ions are only weakly affected, with relative changes typically below 10%. TRANSP-NUBEAM module, gyro-fluid FAR3d code, and hybrid MHD-kinetic CLT-K code have been applied to model sawtooth-induced fast ion transport across a few dedicated and different classes of sawteeth. For sawteeth with fast crash time, the FAR3d and CLT-K simulation results show a reduction in core fast-ion density and primarily in passing fast ion population, which aligns with experimental data. Both experimental and modeling results will be quantitatively compared with Kolesnichenko’s theory [1,2] of sawtooth–fast ion interactions to validate the model.

ACKNOWLEDGEMENTS: This work was supported in part by the US Department of Energy under DE-SC0020337, DE-FC02-04ER54698, DE-FG02-04ER54742

REFERENCE:
[1] Ya. I. Kolesnichenko, V. V. Lutsenko, Yu. V. Yakovenko, G. Kamelander Phys. Plasmas 4, 2544–2554 (1997)
[2] Ya.I. Kolesnichenko, V.V. Lutsenko, R.B. White, Yu.V. Yakovenko Nucl. Fusion 40 1325 (2000)

Speaker: Deyong Liu (General Atomics)

44 EP transport induced by core density collapse in LHD

Speaker: Jacobo Varela Rodriguez (Institute for Fusion Studies UT-austin)

45 Magnetohydrodynamic Stability of ITER-Scale Helical Core Plasmas

Speaker: Dr Panith Adulsiriswad (adulsiriswad.panith@qst.go.jp)

46 Probing Marginally Confined Energetic Ions Using Polarization-Resolved Ion Cyclotron Emission Measurements in KSTAR

Speaker: Mr Seongbin Hong (POSTECH)

47 Runaway electrons during a coil quench in stellarators

Stellarators are among the leading concepts for future fusion reactors, with a key advantage over tokamaks being their practical immunity to large-scale disruptions. The work [1] shows that a rapid shutdown of stellarator coil currents - leading to fast dissipation of poloidal magnetic flux - can nonetheless generate a strong toroidal electric field and drive a runaway-electron avalanche, even without any interruption of the net toroidal plasma current. Simultaneously, the E×B drift displaces these energetic electrons outward. Consequently, the runaway population drifts to the wall and can deposit energy while carrying an appreciable toroidal current. Thanks to the much longer quench timescales the problem is far less serious than in a tokamak.

For parameters representative of Wendelstein 7‑X, it is found that during routine plasma operation the avalanche multiplication remains weak, implying a negligible runaway hazard. The more vulnerable phase is coil ramp-down between discharges, when neutral density is low and collisional friction is reduced; the outcome then depends critically on whether even a small seed population is present. A straightforward preventive measure is to maintain sufficiently high neutral-gas pressure between discharges.

In reactor-scale stellarators, more dangerous runaway generation may occur due to stronger induced fields. Since a radiation-induced seed population is necessarily present in an activated device, an accidental coil ramp-down could convert substantial magnetic energy into wall-damaging runaway currents. Some form of dedicated intervention is therefore likely to be necessary, although there is much more time to mitigate such events than in tokamak disruptions.

[1] Pavel Aleynikov, Per Helander, and H\aa kan M. Smith, Phys. Rev. Applied, Accepted 2 January, 2026 https://doi.org/10.1103/v4s5-p4fr

Speaker: Dr Pavel Aleynikov (Max-Planck-Institut für Plasmaphysik)

48 FIRM3D: Fast ion reduced models in 3D

FIRM3D is an open-source Python/C++/CUDA software suite for modeling energetic particle dynamics in 3D magnetic fields. The core guiding-center integration routines grew out of SIMSOPT (Landreman et al., 2021), but have been extended to include additional physics and diagnostics not typically required in the stellarator optimization context. This standalone framework enables focused development of energetic particle physics capabilities with minimal dependencies, making it accessible to the broader stellarator and plasma physics community. Components of FIRM3D include interfaces with MHD equilibrium and wave stability software (BOOZ_XFORM, AE3D, FAR3D); CPU and GPU parallelized integration of the guiding center orbit equation, with symplectic and Runge-Kutta integrator options; and orbit visualization and transport diagnostics, including Poincaré maps, orbit classification, and weighted Birkhoff averaging. New features, including Coulomb collisions and first wall heat-flux modeling, will be discussed.

Speaker: Elizabeth Paul (Columbia University)

49 Analytical fusion product spectrum calculations including bulk temperature effects in magnetized plasmas

In this work, we show an alternative approach for calculating energy spectra of particles generated via fast ions in magnetized plasmas reacting with a finite temperature bulk. While numerical methods account for all degrees of freedom involved, recently developed analytical methods [1-3] disregard the temperature of the bulk plasma in favor of faster calculations. Here, we still make use of the beam-target approximation for the nuclear reaction, to maintain our model analytical, but retain energy and momentum of the target particles in the energy and momentum conservation laws. This is achieved by using center-of-mass velocity coordinates and noting that the center-of-mass velocity $\bf{v}_\text{CM}$ and the relative velocity $\bf{v}_\text{rel}$ are highly correlated in the gyro-angle coordinate $\Gamma$ describing the azimuthal rotation about the local magnetic field. The model remains analytical, as no Monte Carlo sampling of either of the reactant distributions is required, at the cost of some discrepancy which we quantify in terms of spectral moments relative difference with full Monte Carlo simulations [4,5].

[1] Valentini A. et al 2024 Nuclear Fusion 65 026001
[2] Valentini A. et al 2024 Review of Scientific Instruments 95 083551
[3] Valentini A. et al 2025 Nuclear Fusion 65 046031
[4] Eriksson J., Conroy S., Andersson Sundén E. and Hellesen C. 2016 Computer Physics Communications 199
40–46
[5] Nocente M., Källne J., Salewski M., Tardocchi M. and Gorini G. 2015 Nuclear Fusion 55 123009

Speaker: Andrea Valentini (University of Milano Bicocca)

50 Consistent workflow for evaluating beam-target fusion rate with the orbit-following code ASCOT5

Volumetric neutron sources (VNS) are being developed to produce 14 MeV DT neutrons for fusion materials testing. To limit the size and cost of such VNS, beam-driven systems utilising beam-target fusion reactions are widely studied [1]. Accurate prediction of the beam-target fusion rate is therefore essential for VNS design. Beam-target fusion rates are typically evaluated by first calculating the beam ionisation distribution, then following the ions as they slow down, and finally combining the resulting fast-ion density with the background plasma to obtain the fusion rate. This workflow neglects the depletion of fast ions by fusion reactions during slowing down, causing the steady-state fast-ion distribution—and consequently the fusion rate—to be systematically overestimated.

We implement a self-consistent treatment in ASCOT5 [2] by reducing marker weights at each time step according to the local fusion probability. This accounts for the loss of ions due to fusion during the slowing-down simulation. For a representative VNS spherical tokamak case with 150 keV NBI tritium, neglecting fusion depletion overestimates the fusion power by 0.2% and the steady-state fast-ion density by 0.3%. The latter error propagates into various moments that can be evaluated from the distribution, for instance, the power deposition.

The discrepancy increases with slowing-down time because ions have more opportunity to undergo fusion. In a homogeneous 20 keV plasma with 1 MeV deuterium ions, neglecting fusion depletion overestimates the fusion rate by 1.8%, while the total steady-state fast-ion probability mass differs by 1.5%. These results not only facilitate accounting for the fusion depletion in numerical calculations but also demonstrate the parameter space where such effects could potentially make a significant difference.

[1] C. Bachmann et al., ‘Progress in the concept development of the VNS—a beam-driven tokamak for component testing’, Nucl. Fusion, vol. 66, no. 4, p. 046015, Mar. 2026, doi: 10.1088/1741-4326/ae4e46.

[2] K. Särkimäki, ‘ASCOT5 5.6.3 documentation’. Accessed: May 27, 2026. [Online]. Available: https://ascot4fusion.github.io/ascot5/index.html

Speaker: Joona Sissonen

51 Development of an action-angle quasilinear theory for wave-particle interactions in stellarators

Understanding the interactions between the energetic particle (EP) distribution function with waves is critical for burning plasma operation in future reactor-scale stellarators. We develop a quasilinear theory for nonlinear alpha particle-AE interactions in stellarators, building on a resonance-broadened quasilinear model for tokamaks [1] in action-angle coordinates. Action-angle coordinates are developed for quasisymmetric and omnigenous stellarators. The predicted diffusion of the alpha particle distribution function with respect to particle constants of motion (actions) under the influence of shear Alfvén waves is computed and validated with particle tracing simulations in the open-source FIRM3d code [2], where the diffusion coefficient is learned from data using the SINDy formalism [3].

[1] N.N. Gorelenkov et al 2018 Nucl. Fusion 58(8)
[2] E. Paul et al 2026 arXiv preprint arXiv:2605.16734.
[3] A. Kaptanoglu et al 2022 Journal of Open Source Software, 7(69)

Speaker: Avigdor Veksler (Columbia University)

52 Effects of resistive internal kink mode and fishbone instabilities on alpha particle redistribution and transport coefficients in ITER 15 MA scenario

ITER plans to conduct the D-T fusion experiment in 2039. Effects of the resistive internal kink mode (IK) and fishbone instabilities on the redistribution and the transport coefficients of alpha particles in ITER 15 MA baseline scenario are numerically investigated using the particle tracing code PTC. The results of scanning the perturbation amplitude shows that the IK can result in alphas transporting from regions of $\rho$ < 0.1 to 0.1 < $\rho$ < 0.3, where $\rho$ labels the normalized poloidal flux. Radial redistribution of alphas is greater with increasing the perturbation amplitude. The maximum relative change of the alpha number approaches 60% when the perturbation amplitude rises to 850 G. Simulation results obtained by scanning the fishbone frequency indicate that alphas with different energies are sensitive to the assumed fishbone frequency. Redistribution of alphas with energy lower than 100 keV is the strongest at the fishbone frequency of 5 kHz, while perturbations at frequency of 10 kHz lead to the strongest redistribution of alphas with energies between 100 and 500 keV. Redistributions of alphas by both the IK and fishbone instabilities are all localized in the plasma core region of $\rho$ < 0.3, without causing any particle losses.

For the transport coefficients calculation, results considering the equilibrium field show that, the convection coefficient K distribution is anti-symmetric along pitch, and K value is on the order of 10^0. The diffusion coefficient D is basically zero, indicating no diffusion transport in the absence of perturbations. When perturbation is considered, the K value increases to the order of 10^1. The D value is no longer zero, but the Péclet number Pe $\gg$ 1 in all cases, meaning that the convection of alphas is much greater than the diffusion. Results of scanning perturbation amplitudes and frequencies show that larger amplitude leads to larger value of K (from 10^0 to 10^1), indicating a stronger alphas convective transport. The different perturbation amplitudes does not affect the distribution structure of K, while the change in perturbation frequency has a significant impact on it.

Speaker: Dr Guanming YANG (Southwestern Institute of Physics, Chengdu, China)

53 Frequency-guessing method for reduced modeling of energetic particle modes (EPM)

Computationally inexpensive integrated codes simulating interactions between Alfvén waves and fast ions in tokamak plasmas make use of perturbative models for the intermediate and slow processes of instability growth, saturation, chirping and busting, and associated fast ion transport. Meanwhile, the processes by which an Alfvén mode's spatiotemporal structure forms are assumed to have been completed more rapidly, within the mode's oscillation period, $\tau_0 \equiv 2\pi/\omega_0$. This separation of time scales, together with the assumption of a robust spatial mode structure, underlies the computational efficiency of perturbative models, where the Alfvén mode's time-dependence is reduced to that of a scalar signal $s(t) = A(t)\sin(-\omega_0 t - \phi(t))$ with variable amplitude $A(t)$ and phase $\phi(t)$. For Alfvén eigenmodes residing in gaps between continua, eigensolvers can provide the required accurate input data in the form of the spatial structure $\delta\Phi({\mathbf{x}})$, damping rate $\gamma_{\rm d}$, and initial frequency $\omega_0$. For continuum modes, a reasonable guess for $\delta\Phi(\mathbf{x})$ and $\gamma_{\rm d}$ (if necessary, in frequency-dependent form) could be inferred from the typical fast ion orbit width and continuum structure, but it is difficult to guess the correct seed frequency $\omega_0$. Here, we report results of numerical experiments demonstrating that it is possible to find $\omega_0$ by taking advantage of a prompt frequency shift that was previously discovered with the ORBIT code and interpreted as a product of prompt resonant auto-optimization [1]. Since the terms required for such fast dynamics were truncated in the derivation of the perturbative model, the prompt frequency shift is incomplete but suffices for an iterative optimization scheme.

[1] Bierwage, Duarte & White, Plasma Fus. Res. 16 (2021) 1403087. Link: https://doi.org/10.1585/pfr.16.1403087.

Speaker: Andreas Bierwage (QST)

54 Investigation of electromagnetic waves generated by runaway electrons

Electromagnetic modes are routinely observed in tokamaks during runaway electron (RE) experiments [1-3]. These modes lie in both the Alfvénic and the lower hybrid frequency ranges. Theory and recent experiments suggest that such modes can enhance the pitch-angle scattering of REs [4], help reduce their maximum energy [5] and enhance their radial transport.

These modes therefore play a dual role: they provide a diagnostic tool for the underlying RE distribution function while also offering a means of RE attenuation. The latter can be achieved either via passive self-generation of the waves or via their active external excitation.

This contribution aims at characterizing these modes by introducing a new simulation framework for tokamak RE plasmas, coupling the Fokker-Planck code LUKE [6] with the linear plasma wave solver ALPS [7] – the latter originally developed for astrophysical plasmas and here applied to tokamak parameters for the first time. RE distribution functions are generated by LUKE and passed as input to ALPS, which computes the resulting waves for arbitrary distributions. This LUKE-ALPS framework is then benchmarked against experimental TCV observations, using high frequency electromagnetic wave diagnostics (ICE, LHPI) and hard X-ray diagnostics (BGO, LABrDoRE).

Preliminary results and the current status of this benchmarking effort will be presented, with the aim of establishing a consistent picture connecting the wave signature measured by ICE and LHPI with the hard X-ray spectra measured by LaBrDoRE.

[1] DA Spong, et al. Physical Review Letters, 120(15):155002, 2018.
[2] WW Heidbrink, et al. Plasma Physics and Controlled Fusion, 61(1):014007, 2019.
[3] W Bin, et al. Physical Review Letters, 129(4):045002, 2022.
[4] Tünde Fülöp, et al. Physics of Plasmas, 13(6), 2006.
[5] H Choudhury, et al. Physical Review Letters, 136(2):025101, 2026.
[6] J Decker, et al. Euratom-CEA Report No. EUR-CEA-FC-1736, 2004.
[7] Daniel Verscharen, et al. Journal of Plasma Physics, 84(4):905840403, 2018.

Speaker: Olivier Panico (Ecole Polytechnique Fédérale de Lausanne (EPFL), Swiss Plasma Center (SPC), Lausanne)

55 Isomagnetic transformation for trapped energetic particles under 3D magnetic perturbation

The rigorous evaluation of magnetically trapped particle dynamics in 3D perturbed magnetic fields in toroidal magnetized plasmas remains a fundamental theoretical challenge. Standard bounce-averaging methods are often too rough or produce artificial singularities near the turning points due to the distortion of magnetic surfaces. This mathematical difficulty becomes a crucial physical issue for energetic particles with large trapped-orbit widths and rapid precession frequencies.

To systematically resolve this issue, we propose an Isomagnetic Guiding-Center (IGC) transformation based on noncanonical Lie transformation theory [1,2]. This methodology systematically rectifies a distorted magnetic field into an isomagnetic surface prior to the standard bounce-center transformation [3,4]. Instead of artificially truncating the perturbation, we rigorously encode the perturbed field and displacement into the generator and the gauge function, respectively. By pulling back the dynamics from the isomagnetic space, our formalism systematically isolates the secular radial drift from the bounce-angle-dependent oscillating part without breaking the underlying Hamiltonian structure. Furthermore, the identification of the general bounce-drift resonance and the description of the resultant superbanana orbit are also presented.

References
[1] Robert G. Littlejohn, J. Math. Phys. 23, 742 (1982)
[2] Allen J. Brizard and Taik Soo Hahm, Rev. Mod. Phys. 79, 421 (2007)
[3] Robert G. Littlejohn, Phys. Scr. 1982, 119 (1982)
[4] John R. Cary and Allen J. Brizard, Rev. Mod. Phys. 81, 693 (2009)

Speaker: Prof. Gyungjin Choi

56 Large-bandwidth phase space resolved measurements of fast ion density fluctuations and energy transfer induced by Alfven eigenmodes

Resonant energy exchange between waves and particles is a fundamental concept in plasma physics. It was previously measured in satellite and basic laboratory experiments but has never been measured in the core of a magnetic fusion device. Now, however, thanks to DIII-D’s diagnostic suite, resonant energy exchange between energetic particles and Alfvén waves has been measured. This work establishes a validated experimental methodology for extracting energy transfer rates directly from diagnostic fluctuations without requiring absolute calibration.

The DIII-D imaging neutral particle analyzer (INPA) [1] systems were upgraded with a new set of 16 fast channels, achieving temporal resolution of 350 kHz while maintaining a good resolution in phase space (7 keV in energy and 10 cm in radius). These new capabilities allowed to capture the fluctuations of the confined population of energetic particles both in the Alfvenic frequency range (~100kHz for typical DIII-D parameters) and the tearing and fishbones mode (TM) frequency range (~15 kHz for typical DIII-D discharges). In both frequency ranges, fluctuations of a few percent over the baseline signals are observed.

The measured fluctuation of the confined EP, together with the electric fields extracted from the density and temperature fluctuations [2] allows for the characterization of the energy exchanged between the energetic particles and the waves [3]. The secular energy exchange is mostly carried by the interaction of EP drift velocities with the perpendicular (poloidal) electric field; parallel and radial components are negligible, directly confirming the expected behavior due to large drift of EP orbits. Comparison with ASCOT simulations using a mode structure predicted by the NOVA code agree with the observed energy exchange within the approximation of the calculation.

ASCOT scans show that a magnetic fluctuation amplitude of δB/B ≈ 2×10⁻⁶ (corresponding to ~O(1 eV) temperature fluctuation) produces non-overlapping phase-space islands, which is compatible with coherent INPA signals; a mere ~3× increase causes island overlap and likely diffusive transport, explaining why large-amplitude TAEs (observed in the ramp-up and other scenarios) are invisible to the INPA fast channel.

Speaker: Dr Jose Rueda Rueda (Postdoctoral Researcher)

57 Local gyrokinetic simulations of linear Alfv\'en wave growth in reactor-scale W7-X and Thea Energy’s Helios

The impact of $\alpha$ heating on collisional transport of thermal species and the influence of Alfv\'en waves on fast ion transport are crucial mechanisms driving reactor-scale physics, though both remain important open questions for stellarator fusion power plants (FPPs). To better understand the latter, we compute the linear growth rates of Alfv\'en eigenmodes (AEs) in a reactor scale version of W7-X and in Thea Energy’s Helios equilibrium using fully kinetic linear models in the flux-tube gyrokinetics code GX. Large mode numbers for AEs at reactor scale enable use of the flux tube code for rapid computation of AE growth rates.

Benchmarking against existing flux tube studies of toroidal Alfv\'en eigenmodes in a simple tokamak geometry confirms AE growth can be captured in GX. Extension to W7-X shows growth rates of around $\gamma/\omega_A = 4%$ at an $\alpha$ fraction of 2.5% and a fast ion temperature ratio of $T_{\alpha}/T_i = 350$. Growth rates scale linearly with $\alpha$ density, temperature, and density gradient.

Operational points selected along curves of constant fusion output power of P = 958 MW provide self-consistent profiles for the FPP. A $\beta$-driven transition from AE growth to KBM growth is observed as well as $\beta$-induced frequency gap transitions. AE growth rates of around $\gamma/\omega_A =10%$ are observed at a volume-averaged $\beta$ of 1.9%, before a transition to KBM growth at 2% volume-averaged $\beta$ with growth rates at $\gamma/\omega_A =20%$.

Collisional heating of thermal species induced by $\alpha$ transport is captured using Monte-Carlo simulations of fast ions in firm3d with guiding-center tracing. Global transport solver Trinity updates fusion reactivity profiles handled in firm3d. Nonlinear GX simulations provide heat and particle fluxes of the thermal species to update thermal profiles in Trinity. Preliminary results in the absence of waves are presented.

Speaker: Amelia Chambliss

58 NIMROD MHD simulation of axisymmetric modes observed in the TCV tokamak

Axisymmetric modes (toroidal mode n=0) have been observed in recent experiments on TCV with both co-current and counter-current directions injection of neutral beam [1]. These energetic particle driven global modes, prominently the GAE (Global Alfven Eigenmode), and the VDOM (Vertical Displacement Oscillatory Modes) [2], could potentially be driven by fusion alpha particles in future fusion tokamaks and may play a pivotal role in the stability of those plasmas. Indeed, a recent D-T experiment on JET D-T has demonstrated the excitation of these modes by alpha particles [3]. We have carried out linear simulations with the initial value extended MHD code, NIMROD [4], based on reconstructed experimental profiles from a recent TCV discharge, the shot # 72101. In the first phase of the study, we compared the frequency and the structure of the n=0 modes simulated with NIMROD with those obtained in another study done with the eigen value code, MISHKA, on the same TCV discharge [1]. A satisfactory agreement of the results is found between the two codes for the case of a conducting wall placed at the plasma boundary (ideal case). In the second phase, the advanced feature of NIMROD in simulating the open field line region is utilized to study the case of the actual shape and position of the TCV wall that is separated from the plasma boundary by a low density and temperature halo plasma (real case). The detailed characteristics of the principal modes found in our simulation study will be presented and their relationship with the modes observed in the experiment will be discussed.

[1] Dreval M. et al, Axisymmetric global Alfvén eigenmodes in the TCV tokamak, Nuclear Fusion, 66, 086025 (2026).

[2] Barberis, T., Yolbarsop, A. and Porcelli, F., Vertical displacement oscillatory modes in tokamak plasmas. Journal of Plasma Physics, 88 (5) (2022).

[3] Oliver, H.J.C. et al, Axisymmetric eigenmodes excited by alpha particle energy gradients in JET D-T plasmas, Physical Review Letters, 136, 055101 (2026).

[4] Sovinec, C.R., et al., Nonlinear magnetohydrodynamics simulation using high-order finite elements. Journal of Computational Physics, 195(1): p. 355-386 (2004).

Speaker: DEBABRATA BANERJEE (POLITECNICO DI TORINO)

59 Nonlinear Multi-n TAE Dynamics in KSTAR: Mode Competition, Inter-n Energy Coupling, and Zonal Flow Drive

Most nonlinear simulations of Alfvén Eigenmode (AE) stability study one toroidal mode family at a time, yet in a real tokamak discharge several toroidal mode numbers $n$ are unstable simultaneously. When these families coexist, they compete for the same free energy, couple through three-wave interactions, and collectively drive zonal flows, none of which a single-$n$ model can describe. To explore these effects in a concrete experimental setting, we carried out nonlinear gyrofluid simulations of multi-$n$ Toroidal Alfvén Eigenmode (TAE) dynamics in KSTAR geometry using the FAR3D code, stepping up from a single-$n$ baseline to a full multi-$n$ run to see what changes and why.

Here we presents the nonlinear gyrofluid simulations of Alfvén Eigenmode (AE) stability in KSTAR plasma to investigate the effect of energetic particle (EP) drive on the nonlinear coupling of co-existing Alfvén wave families. All simulations are performed for the KSTAR equilibrium (discharge 21695), with EP beta $\beta_f \approx 0.020$, Lundquist number $S = 5\times10^6$. The single-$n$ baseline, where $n{=}2$ TAE grows ($n{=}2$, with 16 poloidal modes, $m = \pm2$–$\pm9$) gives a clean reference. The $m{=}{-}5/n{=}2$ TAE grows at $\gamma_\mathrm{lin} = 0.075\ \tau_A^{-1}$, saturates around $t \approx 50$–$100 tau_A$, and evolves to $t = 200\;\tau_A$. Extending the basis to include $n{=}1$ shifts dominance entirely to that family, whose saturation energy exceeds $n{=}2$ significantly. Next the simulation pushes further, evolving $n{=}1$, $2$, and $3$ simultaneously (with42 dynamic modes) with $m{=}0/n{=}0$ as a dynamic field.

These simulations reveal how strongly the $n$-families compete. Once $n{=}1$ is present it dominates, effectively suppressing $n{=}2$, unlike any single-$n$ study. In the full multi-$n$ run the dominant linear mode shifts to $m{=}{-}8/n{=}3$ ($\gamma_\mathrm{lin} = 0.059;\tau_A^{-1}$), and at saturation the energy is spread broadly across all three families; which mode dominates depends entirely on which families are simultaneously active. Beyond the mode competition, the multi-$n$ runs reveal genuine inter-$n$ energy transfer through the reduced-MHD Poisson bracket, ${\phi(m_1,n_1),,\phi(m_2,n_2)} \to \phi(m_1{+}m_2,,n_1{+}n_2)$: $n{=}1$ and $n{=}2$ modes drive $n{=}3$ sidebands (upward cascade), while $n{=}2$ and $n{=}{-}2$ drive a $n{=}0$ zonal flow (inverse transfer). The zonal potential $\phi_{00}(r,t)$ builds up from the onset and never decays; it persists at $r/a \approx 0.4$–$0.6$ throughout the entire $t = 0$–$1000;\tau_A$ run, coinciding with the peak of the TAE eigenfunctions. This is not a transient, it looks like a self-sustained zonal state driven continuously by the TAEs. To quantify what is driving this zonal flow we have compute the Reynolds stress (RS) and Maxwell stress (MS) at every timestep using the Wronskian. The analysis of saturation phase shows that the Maxwell stress has a similar shape but smaller magnitude than the Reynolds stress, so it is the electrostatic Reynolds channel that does most of the driving, with the electromagnetic piece playing a supporting role.

The most intriguing observation comes from the frequency-vs-radius spectrograms, computed for each $n$-family and overlaid with the Alfvén continua calculated by ALCON code. Each $n$ mode sits within its own continuum gap at a distinct radial band i.e., lower $n$ near the core, $n{=}2$ slightly outward, $n{=}3$ more localized. As the simulation enters deep saturation, new spectral features appear that were not there at linear onset, power at a higher frequency ($\omega \approx 0.35$–$0.40\;\mathrm{rad}/\tau_A$, above the primary TAE band near $0.20$–$0.25\;\mathrm{rad}/\tau_A$) and at a lower frequency ($\omega \lesssim 0.1\;\mathrm{rad}/\tau_A$). These two features sit near continuum accumulation points in the ALCON overlay, suggesting a three-wave decay of the TAE into a higher-frequency daughter (possibly an EAE or higher-gap EPM) and a lower-frequency mode (possibly a BAE or sub-TAE Alfvénic branch). A definitive mode identification is still under investigation, but the spectral signatures are consistent and reproducible across runs.

These results demonstrate that the nonlinear physics of co-existing multi-$n$ TAEs is not a quantitative correction to single-$n$ results, it is qualitatively different. The dominant mode shifts, the saturation level changes, a long-lived zonal flow appears, and new frequency branches emerge, collectively suggesting that multi-$n$ nonlinear mode coupling must be accounted for in reliable EP transport predictions for KSTAR and next-step burning plasma devices.

Speaker: Dr Haider Rizvi (Korea Institute of Fusion Energy)

60 Nonlinear saturation of Axisymmetric (n=0) modes driven by Energetic Particles in Tokamaks

Axisymmetric (n=0) perturbations in magnetically confined plasmas include distinct global oscillatory modes occurring in the Alfvénic frequency range, such as Global Alfvén Eigenmodes and Vertical Displacement Oscillatory Modes, both of which can couple resonantly to energetic particles and undergo nonlinear saturation in the presence of fast-ion drive. This is particularly relevant in NBI or ICRH-heated plasmas, where fast ion populations provide a source of resonant drive for low-n Alfvénic activity, motivating a quantitative test of whether the observed amplitudes are compatible with energetic-particle-driven nonlinear saturation.

In this work, we develop an analytical framework for the nonlinear saturation of generic n=0 modes by energetic particles. Starting from reduced and hybrid MHD descriptions, we derive the mode structure and the corresponding wave-particle coupling coefficients in a form suitable for the weakly nonlinear theory of marginally driven instabilities. This allows us to extract explicit scaling relations for the nonlinear saturation level, the dependence on equilibrium parameters, and the role of the resonant particle population.

The resulting framework provides predictive expressions for the coupling strength, the saturation amplitude scalings, and the expected order of magnitude of the relative magnetic field perturbation amplitude in the weakly nonlinear regime. These scalings are designed to connect analytic theory with experimentally accessible quantities, enabling direct comparison with observations of axisymmetric fluctuations in present-day tokamaks and offering a route to assess whether their measured saturation levels are consistent with energetic-particle-driven nonlinear dynamics.

Speaker: Simone Cavallero (Polytechnic of Turin, Princeton University)

61 On the drive of axisymmetric modes by fusion alpha particles in tokamaks

There is evidence that axisymmetric modes can be driven by alpha particles in tokamaks [H. J. C. Oliver, Phys. Rev. Lett. \textbf{136}, 055101 (2026)]. It is therefore of interest to investigate the main mechanism behind the drive of the modes. When the distribution function of the non-thermal ions, $f_h$, is given as a function of the invariants of the unperturbed motion energy $\mathcal{E}$, pitch angle variable $\Lambda=\mu B_0/\mathcal{E}$, ($\mu$ is the magnetic momentum) and the toroidal angular momentum $P_\phi$, i.e. $f_h(\mathcal{E},\Lambda,P_\phi)$, the drive of axisymmetric modes requires either a bump on tail in the energy direction (with $\Lambda$ and $P_\phi$ held constant) or an anisotropy characterised by a variation of $f_h$ with $\Lambda$. In the simplest model with only collisional slowing down of alpha particles, $f_h(\mathcal{E},\Lambda,P_\phi)$ always decreases monotonically with $\mathcal{E}$, i.e. there is no drive caused by $\partial f_h / \partial \mathcal{E} |_{\Lambda, P_\phi=const}$. Instead, it is shown that finite orbit width effects lead to an anisotropy with a peak of $\partial f_h/\partial\Lambda$ in the region of trapped orbits. The ability of this anisotropy to drive axisymmetric modes is analysed with a “toy model” where only the slowing down of alpha particles on background electrons is taken into account. The variation of the drive with plasma current and plasma size is investigated.

Speaker: Lars-Göran Eriksson (Chalmers University of Technology, Gothenburg, Sweden)

62 Optimized confinement of alpha particles vs. optimized confinement of NBI ions at high beta in stellarators: a case study of W7-X

Due to the 3D geometry, the toroidal canonical momentum is not an adiabatic invariant in stellarators. Typically, this leads to higher losses of fast ions in comparison with tokamaks. Wendelstein 7-X, an optimized stellarator, is predicted to have improved confinement of isotropically distributed fast ions in equilibria with high beta. Whether this property of the optimized equilibrium can be demonstrated by NBI-generated fast ions, and which parts of the phase space distribution function are actually affected by the optimization – these are the topics of our contribution.

Speaker: Dmitry Moseev (IPP Greifswald)

63 Progress of global gyrokinetic simulations for future fusion devices

We present recent progress of applying global electromagnetic gyrokinetics (GK) to future fusion devices. Results shown from the global GK particle-in-cell (PIC) code ORB5 [E. Lanti et al., CPC, 2020] include simulations of the EUROfusion Volumetric Neutron Source (VNS) [see also Ph. Lauber, this conference], for which simulations of Alfvén eigenmodes have been performed as well as for microinstabilities. Also shown are results of an ITER IMAS scenario #53299/4 [ITPEA-TG].
The closely related 3D-capable global GK PIC code EUTERPE [R. Kleiber et al., CPC, 2024] is also shown for tokamak simulations, and a comparison with the ORB5 results will be presented.
Challenges associated with simulations of large scale devices are discussed, as well as an outlook of planned next-step activities.

Speaker: Thomas Hayward-Schneider (Max Planck Institute for Plasma Physics, Garching)

64 The shear Alfvén continuum in quasi-symmetric and quasi-isodynamic stellarators

Energetic particle (EP)-driven Alfvénic instabilities have been shown to play an important role in stellarators, as demonstrated by neutral beam fast ions in the W7-AS [1] and TJ-II [2] experiments. Alfvénic instabilities may be similarly important in fusion power plants driven by fusion-born alpha particles. However, a comprehensive numerical study of the shear Alfvén continuum in modern stellarator fusion power plant equilibria has not been previously performed. To address this void, we use Stellgap [3] to calculate the shear Alfvén continuum across several quasi-isodynamic and quasi-symmetric configurations — including quasi-helically symmetric and quasi-axisymmetric equilibria. We describe the considerations for selecting poloidal and toroidal modes, along with several newly developed tools, including a method for automatically labeling continuum gaps. Our analysis focuses on frequency ranges resonant with fusion-born alphas or neutral beam fast ions, and we detail how these ranges were determined. The continuum structure of stellarators can differ substantially from that of tokamaks; in particular, we find that stellarator continuum gaps are dominated by HAE gaps with small TAE gaps. We examine the aspects of magnetic geometry that likely drive this gap structure [4]. Finally, we explore possible methods to optimize stellarator equilibria for reduced continuum gap widths in EP-resonant frequency ranges.

[1] A. Weller, et al. Phys. Rev. Lett. 72, 1220–1223 (1994).
[2] R. Jiménez-Gómez, et al. Nucl. Fusion. 51 (3), 033001 (2011).
[3] D. A. Spong, et al. Phys. Plasmas. 10 (8): 3217–3224 (2003).
[4] E. Paul, et al. Journal of Plasma Physics. 91 (4), E101 (2025).

Speaker: Abdullah Hyder (Columbia University)

Physics of alphas and burning plasmas

65 Alpha Particles without DT: Alternative Routes in Future JT-60SA Plasmas

Future fusion power plants will operate in the burning plasma regime, where a substantial fraction of the plasma heating is provided by fusion-born alpha particles produced in DT reactions. Understanding the generation, slowing down and confinement of these energetic particles is therefore essential and remains a major objective of present experimental, theoretical and modelling efforts in preparation for ITER operation.
Following the end of the JET programme, no currently operating large tokamak is able to use DT fuel because of the stringent safety and regulatory constraints associated with the high neutron yield of DT reactions. This is the case for the JT-60SA tokamak in Naka, Japan, presently the largest operating tokamak in the world. Nevertheless, alternative routes for alpha-particle production can be explored in JT-60SA-relevant plasmas ahead of future experimental campaigns.
In this work, first-of-a-kind numerical studies are presented for JT-60SA based on two fusion schemes: D+3He ®4He+p [1] and p+11B®34He [2]. The alpha-particle birth energies associated with these reactions are comparable to 3.5MeV alphas produced in DT fusion: the D+3He channel generates 3.6 MeV alpha particles, while the two most energetic alpha particles produced in the p+11B reaction carry energies of approximately 4 MeV. Since beam–target reaction rates in JT-60SA are enhanced by the high-power (up to 10MW) negative-ion neutral beam injection system [3], capable of beam energies up to 500 keV on both deuterons and protons, the device is a particularly favourable platform for establishing a solid alpha physics programme (see also Fig. 1 from cross section dependency on beam energy in lab frame), along similar lines but at higher alpha birth rates as previously done at JET i.e. investigating alpha generation and confinement, the role of alphas on MHD and turbulence, and the development and validation of dedicated alpha-particle diagnostics in reactor-relevant conditions [4-6,7].
Using the state-of-the-art ASCOT5 suite to model neutral beam injection, fast-ion slowing down and fusion reactions with beam–thermal distributions, we investigate the production and confinement during the typical 0.5- 0.7s of slowing down time of alpha particles arising from both D-3He and p-B scenarios relevant for future JT- 60SA operations. The sensitivity of alpha-particle generation to plasma density, temperature, vertical plasma position is assessed in order to identify trends relevant for scenario optimisation and future experiments. Alpha loss channels from MHD are also investigated. This exploratory work opens the way for a sustained alpha-particle physics programme in JT-60SA. Although the expected fusion yields remain below those achievable in DT plasmas, they are nevertheless expected to be sufficiently high to access key energetic-particle physics processes under reactor-relevant conditions. In particular, recent estimates [8] indicate that, at half the 3He concentration used at JET during the highest alpha birth rates achieved in D+3He experiments (up to 25% 3He concentration), alpha birth rates at least twice as large may be obtained, thereby extending the accessible parameter space for the study of alpha-driven transport, interactions with MHD activity and turbulence, and the possible emergence of collective energetic-particle phenomena not previously observed experimentally.

Figure 1 online
Figure 1: JT-60SA enables studies of alpha-particle physics without introducing tritium into the plasma by utilizing the high-power N-NBI system. (a) Fusion cross-section for D+3He reactions between N-NBI deuterons and 3He ions (in red) and DT reactions between N-NBI deuterons and T ions (dashed black); (b) Fusion cross-section for p-B reactions between N-NBI protons and B impurities (link: https://www.dropbox.com/scl/fi/q0i80ot6u5hs8srjhxk8k/Figure1.png?rlkey=jfe32h06qtftf8bce98j7lrs9&dl=0)

References
[1] Jacquinot, J., Sadler, G. J., Fusion Technology, 21(4), 2254–2264 (1992)
[2] Magee, R.M., Ogawa, K., Tajima, T. et al., Nat Commun 14, 955 (2023)
[3] Garcia J. et al, Nucl. Fusion https://doi.org/10.1088/1741-4326/ae74e1 (2026)
[4] Kazakov Ye. O., et al., Nucl. Fusion 60, 112013 (2020)
[5] Kiptily V., et al., Plasma Phys. Control. Fusion 64, 064001 (2022)
[6] Nocente M., et al., Rev. Sci. Instrum. 93, 093520 (2022)
[7] Kiptily V., see presentation at this conference.
[8] Coelho R., et al., submitted to Nucl. Fusion (2026)

Speaker: Rui Coelho (Instituto Superior Tecnico - IPFN)

66 Aspects of energetic particle physics in future tokamak devices

In next-step fusion devices, the physics of energetic particles (EP) will play a prominent role in distinct ways. For fusion power plants, such as DEMO-type or first-of-a-kind devices [1, 2], the self-organization of an alpha-particle heated plasma strongly depends on the transport and cross- scale coupling properties of the EP-driven meso- and background-driven micro-scale fluctuations mediated by the alpha particle population. In contrast, for neutron technology oriented high neutron flux devices, such as the European volumetric neutron source (VNS) [3], the beam-target physics [4] requires strong external ion injection and high electron heating (e.g. 42.5 MW of neutral beam heating and 8 MW according to the present VNS design point). The resulting anisotropic slowing down distribution dominates the overall properties of the plasma scenario, as it strongly influences the current profile, rotation, and beam-target fusion rate.
In this contribution, the challenges of EP physics in these future devices in relation to present-day experiments and ITER will be discussed. In order to highlight their differences, global quantities such as heating power density, EP pressure fractions and equilibrium distribution functions will be compared. Stability and transport of alpha and beam ions due to Alfvénic modes are investi- gated based on predictive scenario modelling [5, 2]. We use the EP-Stability workflow [6] to scan a wide range of toroidal Alfvén eigenmodes (TAEs), beta-induced Alfvén eigenmodes (BAEs) and reversed shear Alfvén eigenmodes (RSAEs) in an hierarchical way, in order to find stability bound- aries. Realistic NBI sources and distributions as given by BBNBI/ASCOT [13, 14] are used in order to accurately construct the anisotropic distribution functions. In order to assess the related quasi-linear transport, including the possible interaction between the different AE branches, we use the EP transport code ATEP-3D [7, 8, 9] based on the phase space zonal structure transport theory [10, 11, 12].

Authors:
Ph. Lauber$^{1,6}$, M. Falessi$^2$, T. Hayward-Schneider$^1$, V.-A. Popa$^1$, T. Sterkl$^1$, R.Stucchi$^1$, F. Zonca$^2$, C. Bourdelle$^{3,6}$, F. Maviglia$^6$, A. Quartararo$^{6,10}$, M. Siccinio$^1$, P. Zumbolo$^6$, C. De Piccoli$^7$, K. Särkimäki$^5$, A. Snicker$^{4,5}$, P. Vincenzi$^{7,8}$, M. Weiland$^1$,E. Bray$^9$, T. Luda$^1$, E. Fable$^1$, C. Angioni$^1$

$^1$ Max-Planck-Institut für Plasmaphysik, Boltzmannstraße 2, D-85748 Garching, Germany email: philipp.lauber@ipp.mpg.de
$^2$ ENEA, Fusion and Nuclear Safety Department, C. R. Frascati, I-00044 Frascati (Rome), Italy
$^3$ CEA, IRFM, F-13108 St-Paul-Lez-Durance, France
$^4$ Aalto University, Department of Applied Physics, P.O. Box 14100, FI-00076 AALTO, Finland
$^5$ VTT Technical Research Centre of Finland Ltd., Espoo, Finland
$^6$ EUROfusion Consortium, Garching bei München, Germany
$^7$ Consorzio RFX (CNR, ENEA, INFN, University of Padova, Acciaierie Venete), Padova, Italy
$^8$ Institute for Plasma Science and Technology, National Research Council, Padova, Italy
$^9$ NEMO group, Dipartimento Energia, Politecnico di Torino, Italy
$^{10}$ Department of Engineering, University of Palermo, Italy

References:
[1] Coleman M, Zohm H, Bourdelle C, Maviglia F, Pearce A, Siccinio M, Spagnuolo A and Wiesen S 2025 Nuclear Fusion 65 036039
[2] C Bourdelle et al 2026 in preparation
[3] M Siccinio et al 2026 51th EPS Conference on Plasma Physics oral
[4] M Maslov et al 2023 Nuclear Fusion 63 112002
[5] E Bray et al 2026 51th EPS Conference on Plasma Physics oral
[6] Popa V A, Lauber P, Hayward-Schneider T, Schneider M, Hoenen O and Pinches S 2023 Nuclear Fusion 63 126008
[7] Lauber P, Falessi M, Meng G, Hayward-Schneider T, Popa V A, Zonca F and Schneider M 2024 Nuclear Fusion 64 096010
[8] Meng G, Lauber P, Lu Z, Bergmann A and Schneider M 2024 Nuclear Fusion
[9] T Sterkl et al 2026 TMEP Meeting, Lausanne 2026 poster
[10] Falessi M V and Zonca F 2019 Physics of Plasmas 26 022305
[11] Zonca F, Chen L, Falessi M V and Qiu Z 2021 Journal of Physics: Conference Series 1785 012005
[12] Falessi M V, Chen L, Qiu Z and Zonca F 2023 New Journal of Physics 25 123035
[13] Asunta O, Govenius J, Budny R, Gorelenkova M, Tardini G, Kurki-Suonio T, Salmi A and SipilÃC S 2015 Computer Physics Communications 188 33–46
[14] Hirvijoki E, Asunta O, Koskela T, Kurki-Suonio T, Miettunen J, Sipilä S, Snicker A and Äkäslompolo S 2014 Computer Physics Communications 185 1310–1321

Speaker: Philipp Lauber (MPG IPP Garching)

67 Recent progress in modeling and diagnosing energetic particles in the SPARC tokamak

This talk will describe recent progress in modeling and diagnosing energetic particles in the SPARC tokamak, specifically DT-fusion alpha particles and runaway electrons (REs). First, alpha-driven MHD instabilities (such as Toroidal Alfven Eigenmodes) are simulated with several linear and nonlinear hybrid kinetic-MHD codes, including M3D-C1; these show good agreement in predicting a “most unstable” TAE with toroidal mode number n ~ 10. These sub-MHz range fluctuations are expected to be diagnosed via interferometry and arrays of Mirnov coils. Alpha transport - from TAEs, NTMs, and toroidal field ripple - is evaluated via particle following, and the resulting impact on plasma performance, confinement and losses to the first wall are assessed. For expected mode saturation amplitudes dB/B ~ $10^{-3}$, fusion power is predicted to only decrease by ~3%, while alpha power losses to the first wall could increase by a factor of ~2-4. Alpha knock-on reactions, leading to suprathermal fuel ions and high energy neutrons, are also modeled and could be measured via neutron spectrometry. Second, REs and MHD are solved self-consistently with M3D-C1; mitigation via massive gas injection and the passive RE mitigation coil (REMC) are simulated, including both thermal and current quenches, as well as the final RE termination event. REMC-induced stochasticity and resulting RE losses are found to be highly dependent on thermal conductivity and the magnetic geometry, with flux surfaces prone to rehealing near the magnetic axis; for SPARC, the REMC is predicted to be effective, although perhaps not fully preventive. RE transport and impacts on plasma-facing components are similarly evaluated via particle following, with analysis ongoing. REs are anticipated to be diagnosed via their synchrotron emission and hard x-ray bremsstrahlung, among other measurements.

Speaker: Alex Tinguely

68 Alpha particle physics in ARC burning plasmas

The ARC tokamak, currently under design by Commonwealth Fusion Systems, is expected to produce over 1 GW of fusion power in a deeply burning plasma regime with Q > 50.[1] Fusion-produced alphas will carry over 200 MW of power and act as the dominant source of heating in ARC, and if poorly confined, can cause melting or sputtering of the first wall. This work presents an overview of alpha particle physics in ARC.

First, simulations of ripple-induced alpha losses and resulting steady-state heat loads to an axisymmetric first wall were performed using the Monte Carlo orbit-following code ASCOT5, including a scan over TF coil misalignment magnitude, and found 1-2% of alpha power crossing the last closed flux surface (LCFS) with most realistic misalignments. With perfectly aligned TF coils, alpha wall loads were on the order of 100 kW/m^2, concentrated just below the outer midplane.

Next, the effects of MHD instabilities, including toroidicity- and ellipticity-induced Alfvén eigenmodes (TAEs and EAEs), tearing modes, and sawtooth oscillations, on alpha-particle transport were investigated using reduced energetic-particle transport models, such as the Resonance-Broadened Quasilinear (RBQ) model and ORBIT-Kick. These studies predicted generally benign alpha-particle transport and losses. Nevertheless, the cumulative impact of these processes over slowing down timescales may alter the alpha-particle current drive, requiring self-consistent analysis with integrated transport codes such as TRANSP.[2]

Finally, the reduction of ITG turbulence by TAE-induced zonal flows in a reduced plasma current version of ARC was explored using nonlinear gyrokinetic code CGYRO. Significant suppression of ion-scale turbulence was observed near the r/a = 0.35 flux surface in the presence of unstable TAEs, effectively raising the ITG critical gradient by 25%[3]. These effects suggest that in a full-current ARC plasma, we may see performance improvement via this mechanism, as has been predicted for SPARC.[4]

This work is supported by Commonwealth Fusion Systems, and is based on work supported by the U.S. Department of Energy, Office of Science, Fusion Energy Sciences, under the Milestone-Based Fusion Development Program. This work has been carried out within the framework of the EUROfusion Consortium, partially funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200 — EUROfusion). The Swiss contribution to this work has been funded by the Swiss State Secretariat for Education, Research and Innovation (SERI). This research uses resources of the National Energy Research Scientific Computing Center (NERSC), a Department of Energy User Facility.

[1]: J. C. Hillesheim et al. “Overview of the physics basis for the ARC fusion power plant.” J. Plasma Phys. 92 (2026) E69.
[2]: P. J. Bonofiglo et al. “Reduced Transport Modeling of Alpha Particle Physics in ARC Burning Plasmas.” Under review (2026).
[3]: J. Hall et al. “Gyrokinetic simulation of fast ion turbulence stabilization in the ARC tokamak.” In preparation (2026).
[4]: A. Di Siena et al. “How Fusion-Born Alpha Particles Suppress Microturbulence in Burning Plasmas.” Under review (2026).

Speaker: A. LeViness (Commonwealth Fusion Systems)

Wednesday 30 September

EP diagnostics and analysis methods

69 Probing Marginally Confined Energetic Ions Using Polarization-Resolved Ion Cyclotron Emission Measurements in KSTAR

Energetic-ion redistribution and loss induced by MHD activity are critical issues for present fusion experiments and future burning plasmas [1,2]. In future fusion power plants, energetic-particle diagnostics must operate in reactor-relevant environments where access, neutron/gamma radiation, and the use of active beam-based measurements may be limited. Radio-frequency diagnostics are therefore attractive because they can be passive, robust, and potentially compatible with high-radiation environments [3,4]. Ion cyclotron emission (ICE), driven by non-thermal energetic-ion distributions through collective wave–particle interactions near ion cyclotron harmonics, is a promising candidate for diagnosing confined, marginally confined, and loss-related energetic ions [5,6].

In this work, we present recent measurements of ICE in KSTAR neutral-beam-injection-heated plasmas using an in-vessel B-dot probe system developed for high-frequency magnetic fluctuation measurements. High-harmonic ICE bursts were observed and found to exhibit clear temporal modulation. The ICE activity is correlated with sawtooth-like events observed in core ECE measurements, suggesting that the emission is linked to rapid redistribution of energetic ions [8]. Polarization analysis of the measured magnetic fluctuations is performed to constrain the ICE wave mode and propagation properties [3,5]. The measured ICE frequencies are interpreted using the ion cyclotron resonance condition including the Doppler-shift term, allowing an estimate of a quantity related to the fast-ion parallel energy [5,6]. Wave-dispersion analysis is also performed to examine whether the observed ICE is consistent with waves excited by energetic ions redistributed toward barely confined or loss-related orbits near the plasma edge [5,6].

These results suggest that ICE can provide time-resolved information on energetic-ion redistribution beyond simple harmonic identification. In particular, the observed ICE bursts may reflect a transient energetic-ion population that becomes marginally confined or loss-related after sawtooth-induced redistribution. To test this interpretation quantitatively, TRANSP/NUBEAM simulations will be used to reconstruct the fast-ion distribution and orbit characteristics, while FIDASIM will be used to provide synthetic fast-ion diagnostic constraints for comparison with available measurements [7,8]. This work aims to establish ICE as a complementary and reactor-relevant diagnostic for energetic-ion redistribution, marginal confinement, and fast-ion-loss-related physics in KSTAR and future fusion power plants.
This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korea government (MSIT) (Nos. RS-2023-00281272, 2022R1F1A1073863, 2024-00409564, 2026-25536992, 2026-25545529), by the R&D Program of ‘Korea-US Collaboration Research for High Performance Plasma on Tungsten Divertor (EN2603)’ through the Korea Institute of Fusion Energy (KFE) funded by MSIT, and by Glocal University 30 project. The authors acknowledge Dr. Andreas Bierwage for helpful discussions on energetic-ion physics and ICE interpretation.

[1] M. H. Kim et al., Nucl. Fusion 58, 096034 (2018).
[2] M. H. Kim et al., Nucl. Fusion 60, 126021 (2020).
[3] R. Ochoukov et al., Rev. Sci. Instrum. 86, 115112 (2015).
[4] B. C. G. Reman et al., Nucl. Fusion 59, 096013 (2019).
[5] B. Chapman et al., Plasma Phys. Control. Fusion 62, 055003 (2020).
[6] N. N. Gorelenkov, Plasma Phys. Rep. 42, 430 (2016).
[7] B. Geiger et al., Plasma Phys. Control. Fusion 62, 105008 (2020).
[8] X. D. Du et al., Nucl. Fusion 60, 112001 (2020).

Speaker: Mr Seongbin Hong (POSTECH)

70 Investigation of Fast-Ion Velocity Distribution Functions via Neutron Emission Spectroscopy on EAST

This report presents an overview of the experimental results and data processing advancements in fast-ion diagnostics using neutron emission spectroscopy (NES) on the EAST tokamak.

The EAST NES system comprises multiple spectrometers arranged along four lines of sight (LOS), including a time-of-flight enhanced diagnostic (TOFED), a stilbene crystal, three liquid scintillators, a CLYC, a LaCl$_3$ spectrometer, and a newly installed single-crystal diamond spectrometer in the 2026 campaign. These detectors, varying in viewing angles and physical principles, cover different regions of the fast-ion phase space (particularly energies > 100 keV), enabling a systematic evaluation of diagnostic capabilities.

During synergistic NBI and third-harmonic ICRF heating, a significant broadening of the neutron spectra was observed across multiple spectrometers, indicating fast-ion acceleration well beyond the injection energy [1]. TRANSP simulations, validated by synthetic spectra generated by the GENESIS code and the instrument response functions (IRFs), as well as a weight-function analysis, revealed a fast-ion tail extending up to ~600 keV with a pronounced pitch-angle asymmetry favoring co-current velocities. Orbit calculations demonstrated that this asymmetry is primarily governed by the fast-ion loss boundaries on EAST.

Furthermore, in high poloidal beta ($\beta_{\mathrm{p}}$) discharges that achieved a record fusion neutron rate ($S_{\mathrm{n}}=3.9\times 10^{14}\,\mathrm{s}^{-1}$) on EAST, fast-ion behaviors under NBI and second-harmonic ICRF heating were analyzed by combining NES and neutron flux monitors (NFMs) [2]. Analysis shows that while the synergistic effect directly contributes $\sim30\%$ to the total neutron rate via the formation of a high-energy fast-ion tail, this enhancement is partially offset by NBI-induced profile degradation. Orbit phase-space analysis using the ORBIT code further revealed that synergistic heating drives suprathermal ions into smaller orbits (e.g., stagnation orbits), leading to the spatial redistribution of fast ions and the peaking of the neutron emissivity profile.

By systematically evaluating detectors across various distances, shielding configurations, and detection principles (TOF, organic, and Cl-based inorganic scintillators), this work establishes a comprehensive benchmark of NES for fast-ion diagnostics. Ongoing efforts focus on 4-LOS NES-only tomography and the development of orbit-space weight functions to understand the pronounced orbit effects typical of medium-sized tokamaks like EAST. Collectively, these advancements provide critical insights into high-fidelity fast-ion measurements, offering a robust technical framework for the design and optimization of NES-based diagnostics in future burning plasma devices.

References
[1] A.D. Xu et al 2026 Nucl. Fusion 66 066024
[2] A.D. Xu et al 2026 Nucl. Fusion (submitted)

Speaker: Andong Xu (Peking University)

71 On confined alpha-particle studies in low-activation plasmas

Studying the mechanisms of energetic α-particle slowing down, redistribution and the development of optimal plasma conditions for their confinement is a priority task for burning fusion reactors. The harsh radiation environment of deuterium-tritium (D-T) fusion reactors makes realisation of this task extremely difficult since a restricted set of plasma diagnostics will be available. In this presentation, we discuss various low-activation plasma scenarios and diagnostics for existing and forthcoming fusion machines that could allow α-particles study to be studied prior to reactor plasmas.
In JET, confined α-particles were studied in several low-activation plasma scenarios, generating 3.6-MeV alphas with the D(3He,p)4He fusion reaction, i.e. using the following ICRF plasma heating schemes: 1) 3He-minority ICRF heating of D-plasmas; 2) D-ion beam acceleration by 3rd harmonic ICRH and in the 3-ion ICRH scheme in D-3He plasmas. Also, confined MeV α-particles were studied in He-plasmas by accelerating the 4He-ion beam with 3rd harmonic ICRF and generating D-T α-particles with short T-NBI blips in deuterium plasmas. The confined α-particle diagnosis was based on detection of γ-rays from the nuclear reaction 9Be(α,nγ)12C [1].
In addition to the D(3He,p)4He reaction, there is another aneutronic fusion reaction, 11B(p,2α)4He, which also produces MeV α-particles and can be used for studies [2]. To generate highest density of α-particles with these fusion reactions, auxiliary heating with 3He-, D- and H-ions, as well as boron impurity are needed. Specifically, for α-particle studies we propose plasma heating scenarios with energetic hydrogen and deuterium neutral beam injection and ICRF heating.
Gamma-ray and neutron diagnostics, which are amongst a restricted set of reactor-compatible diagnostics, can be considered for confined α-particle measurements and characterisation of the fast H-, D- and 3He- ions in these low activation plasmas. For this purpose, nuclear reactions generating the required γ-rays and neutrons are selected [3]. Modelling and assessments of γ-ray and neutron diagnostic reactions show that a comprehensive study of α-particles could be performed in a currently working fusion devices, i.e. JT-60SA [4,5], prior to the operation of high-performance D-T plasmas.

1. V.G. Kiptily, F.E. Cecil and S.S. Medley, 2006 Plasma Phys. Control. Fusion 48 R59
2. K. Ogawa et al 2024 Nucl. Fusion 64 096028
3. V.G. Kiptily et al 2026 Nucl. Fusion 66 066004
4. Ye.O. Kazakov et al., 30th IAEA Fusion Energy Conference (2025)
5. R. Coelho, see presentation at this meeting

Speaker: Dr Vasili Kiptily (UKAEA, United Kingdom)

72 Qualitative validation of CAD-based simulations of NBI-induced prompt fast-ion losses against IRTV measurements in KSTAR

Modern devices such as KSTAR and ITER are crowded with heating ports, cooling channels, and diagnostics, and rely on auxiliary heating such as neutral beam injection (NBI), so that beam power deposition and the resulting plasma-facing component (PFC) heat loads are governed by real, non-axisymmetric engineering geometry that idealized toroidally symmetric models cannot capture. CAD-to-simulation tools now embed this engineering detail by tracing field lines or ion orbits through 3D CAD geometry to compute power deposition and PFC heat flux (e.g., SMARDDA; the HEAT toolkit [1, 2]), and are being assembled into AI-augmented digital twins (e.g., DIII-D and MAST-U in NVIDIA Omniverse [3, 4]), moving the field toward an AI-native treatment of engineering detail. Despite this progress, experimental validation that such CAD-based predictions reproduce the measured 3D global heat-load footprints — including the localized peaks that arise at leading edges and limiters — remains limited, yet is a prerequisite for trusting these tools in design and between-shot operation.
We report a qualitative validation of Monte Carlo orbit following code NuBDeC [5], a modular and unified CAD-to-simulation framework that performs high-performance fast-ion–wall collision detection on CAD-derived unstructured meshes and maps the per-face heat flux deposited by NBI-induced prompt fast-ion losses onto the 3D PFC surface, with a CAD categorization that links every mesh face to its engineering component (divertor, poloidal limiter, passive stabilizer). Because the pipeline operates on the 3D geometry with engineering details, it predicts non-axisymmetric, locally peaked loss footprints at leading edges and limiters. To test these predictions, we conducted dedicated KSTAR experiments using infrared thermography (IRTV), following the W7-X fast-ion IR validation precedent [6]. Discharges were run in a diverted L-mode (lower single null) held below the L–H transition to suppress ELM- and thermal-ion-driven signals and to steer lost ions into the IRTV field of view; lower Ip and BT were used to enhance the toroidal drift and the prompt-loss signal [5].
Across dedicated shots scanning plasma current (Ip = 400 / 500 / 600 kA), toroidal field (BT = 1.7 / 1.8 T), and beam sources (NB1C, NB2A/B/C), L-mode flat-tops were obtained. Fitted ne and Te profiles from Thomson scattering diagnostics were supplied to NuBDeC, and the simulated prompt-loss heat-flux maps were compared against the beam-on/off IRTV ΔT maps accumulated over 3–7 s window for each shot. For NB1C at fixed BT and power, lowering Ip from 600 to 400 kA enhanced the predicted prompt loss, and the corresponding IRTV ΔT difference shows the localized temperature rise appearing at the predicted leading-edge and poloidal-limiter locations — qualitatively consistent with simulation. The comparison is made under the following assumptions: thermal-ion losses are confined to the divertor region and therefore fall outside the predicted prompt-loss footprints, and the delayed orbit-loss distribution deviates only slightly from the prompt-loss distribution. This first qualitative benchmark confirms that the CAD-based pipeline reproduces where, and how, NBI prompt-loss heat loads concentrate on KSTAR PFCs, establishing a standardized, error-resistant foundation for synthetic diagnostics.
Building on this validated pipeline, we are planning to extend NuBDeC toward an AI-native treatment of engineering detail. The surrogate model can be the future research goal that predicts the prompt-loss heat flux directly on the unstructured PFC mesh, conditioned on the magnetic equilibrium, the beam source, and the as-built geometry. By preserving the full axial asymmetry and engineering-level detail while collapsing per-case cost, the surrogate aims to make exhaustive NBI parameter scans practical during conceptual design of future devices and to deliver between-shot heat-load guidance that helps avoid wasted discharges and fine-tune heating to each experiment’s mission.

Speaker: Taeuk Moon (Korea Institute of Fusion Energy)

10:50 Coffee break

Physics of EP modes and transport

73 A Diffusive Electron Cyclotron Emission Fast Ion Transport Model

An empirical method for inferring MHD-induced anomalous fast ion transport from normalized Electron Cyclotron Emission (ECE) temperature fluctuation measurements is reported. The model developed infers a time-dependent, spatially varying anomalous fast ion diffusion profile directly from normalized temperature fluctuations such that $D_{fast} \propto dT/T$. The model is coupled to the Monte Carlo fast ion code NUBEAM within the plasma transport code TRANSP to simulate DIII-D plasmas, and found to reproduce experimental measurements from five different fast ion related diagnostics under strong Alfvén Eigenmode activity conditions: the volumetric neutron rate, the plasma stored energy, spatially resolved Main Ion and Fast Ion D-𝛼 spectra, and phase space resolved Imaging Neutral Particle Analyzer (INPA) measurements. The application of this method to a database of more than 40 DIII-D discharges, spanning a large fraction of the available DIII-D parameter space, finds the predicted neutron rates and stored energies agree with experiment with an average accuracy of 15% or better. Detailed analysis of select simulations using the D(ECE) fast ion transport model indicate critical-gradient-like behavior, with direct comparisons of D(ECE) predictions to the reduced critical gradient model TGLF-EP+Alpha finding strong agreement. Cross-comparison of empirical transport predictions against a suite of reduced fast ion transport modeling results from the codes KICK, RBQ, and TGLF-EP+Alpha finds the strongest agreement with TGLF-EP modeling, highlighting the efficacy of a local critical gradient fast ion transport modeling framework. This empirical modelling work demonstrates the efficacy and methodology behind a new standard tool for routine analysis of EP experiments which is accessible to non-experts, contains no “tuned” free parameters, and has trivial computational cost.

Speaker: Dr Kyle Callahan (UCI)

74 Maximising Beam-Plasma Fusion Rate in 1 MA MAST-U Discharges

The MAST-U spherical tokamak [1] with aspect ratio R/a [m] = 0.8/0.5, max B=0.72 T at 0.8 m, elongation up to 2.5, operated at 1 MA currents recent two years and delivered high D-D fusion beam-plasma performance. This performance was achieved with two deuterium NBI sources, one of which was on-axis, and the other - off-axis (shifted by 65 cm from the magnetic axis up). Of particular interest for maximising the neutron yield was the scenario where the beam-driven fishbones and TAEs could be suppressed by combining on-axis and off-axis beams with opposite radial gradients at flat/ weakly non-monotonic central q-profile. A quiescent TAE-free time window was established in such scenario, and the beam-plasma neutron rate was maximised then. The sequence of MAST-U discharges with such TAE-free time window at the end of every discharge, was performed with and without the uplift of the plasma, and the neutron rates, which started from the reference ~10^14 s-1, then achieved ~2.5x10^14 s-1, ~2.8x10^14 s-1, and ~3x10^14 s-1.
Plasmas in these discharges were of the hot ion type, with ion temperatures ~3 keV and electron temperatures ~1.5 keV, with a hollow electron density profile and hollow toroidal rotation profile. Internal Reconnection Events (IREs) [2, 3] appeared to be common in such MAST-U plasmas and these exhibited many properties similar to the disruptions: rapid increase in plasma current, negative spike in loop voltage, increase in elongation, and reduction in plasma energy. However, no terminations of the current were caused by IREs. The plasma density was limited by ~4x10^19 m-3 making the beam slowing-down time long enough for high yield beam-plasma fusion. Similar IREs were recently investigated with neural networks on the ST-40 tokamak concluding that IREs were less likely at higher q(95%) and β_pol. [3]. Based on this knowledge, we proposed and tested a “current ramp-down” technique used just before an IRE. Such technique did mitigate the IRE successfully followed by an increase in plasma density up to ~8x10^19 m-3.
[1] J.R. Harrison et al., 2019 Nucl. Fusion 59 112011; [2] R. Buttery et al., 1996, Proceed. of 23rd EPS, Kiev, Part I, p.416; [3] I. Semenov et al., 2003 Phys. of Plasmas 10 664; [4] C. Windsor et al., 2026, (to be submitted for publication).

Speaker: Sergei Sharapov (UK Atomic Energy Authority, UK)

75 Radial electric field induced by NBI fast ions

Fast neutral beam injection into magnetically confined plasmas produces pairs of fast ions and electrons at their ionization location. While the electrons remain close to the ionization flux surface, fast ions experience substantially larger drift orbits, resulting in a radial displacement between positive and negative charges. This charge separation contributes to the formation of a radial electric field Er [1]. Additional charge separation arises from fast ion losses. Fast ions born inside the confined plasma region may be lost to plasma-facing components, leaving the associated electrons confined within the plasma. Furthermore, charge-exchange (CX) fast ion losses typically occur at larger radii than the original fast ion ionization locations, producing thermal ions and additional radial redistribution of charge. The combined effect of these mechanisms leads to the generation of an NBI-induced radial electric field.
Fast ion ionization sources, distributions, and losses are evaluated using the FIDASIM [2] and EBdyna [3] codes. The resulting radial electric field is estimated using a charge-separation based formulation. Since the radial electric field modifies fast ion trajectories through orbit squeezing effects [4], the calculated Er is subsequently included in a new EBdyna simulation and the radial electric field is re-evaluated. This iterative procedure is repeated several times in order to assess the self-consistent evolution of the fast-ion-induced radial electric field.
The analysis focuses on a QH-mode discharge in the DIII-D tokamak, where plasma access is governed by the E × B shear threshold [5]. Charge-exchange losses are evaluated using a two-dimensional neutral density distribution obtained from DEGAS2 [6] simulations constrained by experimental measurements of background neutrals.

References
[1] Xingyuan Xu, Yingfeng Xu, Xiaodong Zhang, and Youjun Hu. Simulations of the radial electric field induced by neutral beam injection in a tokamak. Nuclear Fusion, 61(8):086002, jun 2021.
[2] Benedikt Geiger, Luke Stagner, William W Heidbrink, Ralph Dux, Rainer Fischer, Yutaka Fujiwara, Alvin Garcia, Asger Schou Jacobsen, Anton Jansen vanVuuren, Alexander N Karpushov, Deyong Liu, Philip Adrian Schneider, Igor Sfiligoi, Peter Zsolt Poloskei, and Markus Weiland. Progress in modelling fast-ion d-alpha spectra and neutral particle analyzer fluxes using fidasim. Plasma Physics and Controlled Fusion, 2020.
[3] F. Jaulmes, G. Zadvitskiy, K. Bogar, M. Imrisek, J. Hromadka, S.Y. Cats, J. Varju, M. Komm, and R. Panek.Modelling of charge-exchange induced nbi losses in the compass upgrade tokamak. Nuclear Fusion, 61(4):046012, 2021.
[4] G.J. Kramer, A. Bortolon, A. Diallo, and R. Maingi. The formation of an radial edge electric field due to finite ion orbit width effects is the possible root cause of the h-mode edge. Nuclear Fusion, 64(10):106035, 2024.
[5] TM Wilks, AM Garofalo, PH Diamond, ZB Guo, JW Hughes, KH Burrell, and Xi Chen. Scaling trends of the critical e× b shear for edge harmonic oscillation onset in diii-d quiescent h-mode plasmas. Nuclear Fusion,
58(11):112002, 2018.
[6] Daren Stotler and Charles Karney. Neutral gas transport modeling with degas 2. Contributions to Plasma Physics, 34(2-3):392–397, 1994.

Acknowledgement
This work has been carried out within the framework of the EUROfusion Consortium,funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No 101052200 — EUROfusion). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Com-
mission. Neither the European Union nor the European Commission can be held responsible for them. Work supported by US DOE under DE-SC0020337 and DE-FC02- 04ER54698.

Speaker: Klara Bogar (Institute of Plasma Physics of the CAS, Prague, Czech Republic)

12:45 TCV visit (please indicate your interest when submitting your Registration Form, or visit https://tmep26.epfl.ch/visit-to-tcv/). Free afternoon.

Thursday 1 October

Synergy EP and turbulence

76 Energetic-particle-driven Alfvénic bursts and zonal structures in JET: DBS measurements and gyrokinetic simulations

We report Doppler backscattering (DBS) measurements and gyrokinetic simulations of alternating Alfvénic bursts in a JET plasma previously analysed by Ruiz Ruiz et al. [Phys. Rev. Lett. 134, 095103 (2025)]. The DBS measurements show that fluctuation power in the ITG-turbulence frequency range is anticorrelated with the fluctuation power of modes in the gaps of the Alfvénic spectrum, with the ITG-range fluctuation power decreasing by nearly an order of magnitude during the Alfvénic bursts. The simultaneous growth of the fundamental TAE, higher-frequency sidebands, and ITG turbulence suppression indicates that the dynamics cannot be described by a simple delayed predator-prey cycle alone. The second sideband grows at approximately twice the rate of the fundamental TAE, consistent with beat-driven generation. Linear gyrokinetic stability analysis with CGYRO shows that the low-wavenumber EP-driven mode is localized near TAE half-rational surfaces, where neighbouring poloidal harmonics overlap. Nonlinear gyrokinetic simulations reproduce alternating Alfvénic bursts on timescales comparable to the DBS observations and show nonlinear generation of zonal fields and structures. Although zonal flows are generated near the half-rational surfaces of the fundamental TAE, their shearing rate remains below the ITG linear growth rate, suggesting that additional mechanisms contribute to the observed turbulence suppression. The simulations also show flattening of the energetic-particle gradient and of the thermal density and temperature gradients near the half-rational surfaces of the dominant TAE modes. These results suggest that zonal density and temperature structures contribute to TAE saturation and to the associated stabilization of ITG turbulence.

Speaker: Dr Juan Ruiz ruiz (University of York (EU))

77 Experimental observation of microturbulence suppression induced by Alfvén Eigenmodes in DIII-D tokamak

The suppression of ion temperature gradient (ITG) turbulence by the nonlinear evolution of toroidicity-induced Alfvén eigenmodes (TAEs) is demonstrated in the DIII-D tokamak [1]. Systematic measurements of plasma parameters reveal that this phenomenon is characterized by: (1) a reduction in the poloidal correlation length of ITG from 5 cm to 2 cm and a 30% decrease in the radial correlation length, coinciding with the formation of a local transport barrier; (2) the emergence of a narrow shear flow layer with a shearing rate exceeding the ITG decorrelation rate; (3) increased Reynolds stress in electron diamagnetic drift direction, with a radial broadening of ~2 cm, which is correlated to the significantly contracted TAE radial mode structure; (4) a deviation of TAE polarization from the conventional shear Alfvén wave towards a more electrostatic polarization by 15%. (5) a localized change in the safety factor during the ITG suppression phase, measured by Motional Stark Effect diagnostics and (6) imaging neutral particle analyzer tomography indicating an increased fast-ion content during ITG suppression, despite the increased total amplitudes of TAEs. Furthermore, a series of the experiments reveals that TAEs routinely suppress ITG turbulence in plasma with higher local safety factor, elevated fast ion beta and larger population of fast ions on passing orbits, i.e., conditions consistent with high linear growth rate of TAEs. Database analysis further identifies the existence of a threshold in TAE drive required for ITG suppression.

Work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences,
using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Award(s) DE-FC02-04ER54698,DE-SC0020337, DE-FG02-08ER54999, DE-SC0020287, DESC0015878, DE-FG02-97ER54415, DE-SC0014664.

[1] X.D. Du et al., Phys. Rev. Lett. 135, 265101 (2025).

Speaker: Xiaodi Du (GAT - Fusion Energy Research General Atomics (US))

78 On How Zonal Fields Suppress Reversed Shear Alfvén Eigenmode in Tokamak Plasmas: Simulation and Theory

Achieving good confinement in tokamak advanced scenarios requires understanding and controlling Alfvénic instabilities driven by energetic particles (EPs). Reversed-shear Alfvén eigenmodes (RSAEs) [1, 2] are commonly observed in plasmas with nonmonotonic safety-factor profiles and can cause substantial EP transport and losses, making their nonlinear saturation mechanisms critical for predicting burning-plasma performance [3]. Zonal electromagnetic fields (ZFs)—toroidally symmetric, low-frequency perturbations of the equilibrium—are a natural candidate to mediate saturation because they arise from self-consistent nonlinearities and can couple scales across the plasma [4-6]. Previous theoretical work emphasized zonal-flow effects and continuum damping as key saturation channels [7, 8], while recent large-scale simulations and analyses have shown both thermal-plasma-driven ZFs and phase-space zonal structures (PSZS) can be important, and in some cases ZFs may even enhance EP drive [9-12]. This motivates a first-principles clarification of which nonlinear channel dominates and how it leads to saturation.

In this work we isolate the thermal-plasma nonlinearity route by keeping EP dynamics linear. Global gyrokinetic simulations (GTC) [13] of a single-n RSAE show clear downward frequency chirping, excitation of radially propagating kinetic Alfvén waves (KAWs), and eventual saturation at finite amplitude, even with linear EP drive. To interpret these results we derive two compact eigenmode models. The “RSAE-ZF-MHD” model (based on the ideal-MHD continuum) demonstrates that beat-driven zonal current produces a dominant downward frequency shift (zonal flow yields a smaller opposite effect) and that continuum resonant damping is negligible. The extended “RSAE-ZF-KAW” model, which includes finite-ion-Larmor-radius terms and electron Landau kinetics, captures the resulting mode conversion to KAWs and the associated strong radiative damping responsible for mode suppression and saturation. Numerical solutions of these models quantitatively reproduce the gyrokinetic observations. These results establish zonal-current-induced frequency chirping plus KAW radiative damping as an efficient saturation mechanism for RSAEs.
References
[1] H. Kimura, et al., Nucl. Fusion 38, 1303 (1998).
[2] S. E. Sharapov, et al., Phys. Lett. A 289, 127 (2001).
[3] P. Liu, et al., Phys. Rev. Lett. 128, 185001 (2022).
[4] L. Chen, et al., Nucl. Fusion 41, 747 (2001).
[5] P. H. Diamond, et al., Plasma Phys. Control. Fusion 47, R35 (2005).
[6] L. Chen and F. Zonca, Phys. Rev. Lett. 109, 145002 (2012).
[7] S. Wei, et al., J. Plasma Phys. 87, 905870505 (2021).
[8] T. Wang, et al., Plasma Sci. Technol. 26, 053001 (2024).
[9] F. Zonca, et al., New J. Phys. 17, 013052 (2015).
[10] F. Zonca, et al., Journal of Physics: Conference Series 1785, 012005 (2021).
[11] M. V. Falessi,et al., New J.Phys. 25, 123035 (2023).
[12] L. Chen, et al., Nucl. Fusion 65, 016018 (2025).
[13] Z. Lin, et al., Science 281, 3 (1998).
[14] J. Bao, et al., Nucl. Fusion 63, 076021 (2023).

Speaker: Ruirui Ma

79 Properties of Alfven eigenmodes during strong turbulence suppression*

Alfvén eigenmodes (AE) create a regime in the DIII-D tokamak where zonal flows and currents strongly suppress ion-temperature-gradient (ITG) turbulence. This oral focuses on the differences between AEs in the strongly suppressed ITG regime and “normal" AEs in unsuppressed plasmas. Strongly suppressed cases are compared to closely matched reference shots at several stages: (1) before the relatively gradual forward transition to the suppressed state, (2) early in the suppressed state, (3) deep in the suppressed state, and (4) after the sudden “back" transition to a normal state that has ITG turbulence and lower temperature. During the strong ITG suppression, AE amplitudes increase and radial eigenfunctions shrink; the number of unstable modes also increases. The primary toroidal and poloidal mode numbers remain unchanged in the suppressed state, but changes in poloidal mode composition are observed. The radial phase shift (rotational twist) of the eigenfunction changes from convex to concave, indicating a reversal of energy flow in the suppressed region. The TAEs burst intermittently with similar statistical properties in both normal and suppressed plasmas. The ITG suppressed state is triggered when one or more AEs acquire an electrostatic component to their polarization; other AEs often acquire an electrostatic component later in the suppressed state. Preliminary analysis suggests that polarization changes occur when the ratio of AE power to mode width exceeds a threshold. Speculation about the causes of these effects conclude the talk.

*Work supported by the U.S. Department of Energy, Office of Science, Office of Fusion Energy Sciences, using the DIII-D National Fusion Facility, a DOE Office of Science user facility, under Award(s) DE-FC02-04ER54698, DE-SC0020337, DE-FG02-08ER54999, DE-SC0020287, DESC0015878, DE-FG02-97ER54415, DE-SC0014664.

Speaker: William Heidbrink

11:00 Coffee break

Non-linear EP physics

80 Phase-space mechanisms of ICRF-driven bursting shear Alfvén waves and their suppression via phase-space engineering

Recurrent bursting Alfvénic instabilities can severely degrade energetic-particle (EP) confinement and impose transient loads on plasma-facing components. We recently reported the first comprehensive simulations of ICRF-heated tokamak plasmas on the slowing-down timescale with the kinetic-MHD hybrid code MEGA, where shear Alfvén wave induced minority ion transport is self-consistently retained during high-energy-tail formation [1]. Both bursting and non-bursting toroidal Alfvén eigenmodes (TAEs) were obtained: on-axis or inboard ICRF resonance yields bursting TAEs, which appear as a set of radially separated discrete modes at distinct, nearly fixed frequencies without chirping, whereas outboard heating always produces a single broad structure whose harmonics share nearly the same frequency, independent of RF power.

This work clarifies the mechanism through EP phase-space analysis. The deeply trapped ICRF-accelerated minority ions resonant with a fixed-frequency TAE occupy a highly localized region of ($E$, $\mu$) phase space. For on-axis/inboard heating, TAE transport carries these ions away from the RF layer, so replenishment is weak. The resulting discrete TAEs keep the resonances isolated, free energy accumulates, and once the growing amplitudes force neighboring resonance islands to overlap, a burst is triggered, consistent with the resonance-overlap (Chirikov) avalanche picture [2]. For outboard heating, the resonant ions stay close to the RF layer, where the strong velocity-space diffusion sharply raises the replenishment rate νeff and restores them faster than the wave flattens the local gradient, suppressing isolated phase-space structures. This high-νeff picture also explains why the non-bursting state sustains a higher EP beta than the bursting one, which a critical-gradient argument alone cannot.

Building on this mechanism, we report for the first time simulations in which an on-axis-driven bursting state is suppressed by an additional low-field-side ICRF wave. The auxiliary wave introduces velocity-space diffusion where the burst-transported ions accumulate, driving the resonant population into the high-νeff regime and mitigating the bursts. To our knowledge this is the first comprehensive simulation to demonstrate phase-space engineering of energetic ions [3] for the control of bursting Alfvénic activity in magnetic confinement fusion.

Acknowledgments: This work was supported by JSPS KAKENHI with Grant Number JP21H04973, JP24K17032 and JP24H00207, the NINS program of Promoting Research by Networking among Institutions (Grant Number 01422301).

References
[1] J. Wang et al., Nucl. Fusion (2026), accepted (doi:10.1088/1741-4326/ae72a4).
[2] H.L. Berk et al., Nucl. Fusion 35, 1661(1995).
[3] J.P. Graves et al., Nat. Commun. 3, 624 (2012).

Speaker: Jialei Wang (National Institute for Fusion Science)

81 Self saturation of Alfén eigenmodes with a reactor-like low ρ*

A new self-saturation regime for energetic particle (EP) driven Alfvén eigenmodes (AEs) at reactor-like ρ is identified using the CGYRO [1] gyrokinetic flux-tube code. This new saturation regime could markedly reduce the predicted impact of AEs in future reactors. Previous nonlinear gyrokinetic flux-tube studies [2,3] of fast ion-driven AEs at higher values of ρ associated with modern experiments have shown highly stiff transport behavior, generally manifesting as a failure to saturate above a critical energetic particle EP driving gradient. This stiff behavior has served as the basis for the TGLF-EP+Alpha model of EP transport and profile prediction [4]. In modern machines, the AE spectrum consists of only a few modes, around five distinct values of toroidal n numbers. In a reactor, the spectrum is much denser, with AEs dominant at the lowest ~30 modes. The present CGYRO simulations, performed at various radii within the Holland compact reactor use case [5] steady-state scenario, show that self saturation can occur with modest diffusivities at EP gradient drive strength well above the critical gradient predicted by the TGLF-EP model. The predicted transport converges only with mode spacing of ∆𝑛 ≤ 2, a so-called half-torus simulation. Saturated diffusivities are found to mostly drop with ∆𝑛 as the AE spectrum becomes denser. We hypothesize that the increased nonlinear n-to-n AE coupling provides the needed saturation mechanism. We propose a finite-stiffness correction to the TGLF-EP+Alpha model for low-ρ* cases consistent with these results that predicts significantly less fusion performance degradation than the current model.
[1] J. Candy and E.A. Belli. Spectral treatment of gyrokinetic shear flow. J. Comput. Phys., 356:448, 2018. doi:10.1016/j.jcp.2017.12.020.
[2] E.M. Bass and R.E. Waltz, Phys. Plasmas 17, 112319 (2010). https://doi.org/10.1063/1.3509106
[3] E. M. Bass and R. E. Waltz Phys. Plasmas 24, 122302 (2017) https://doi.org/10.1063/1.4998420
[4] He Sheng, R. E. Waltz, and G.M. Staebler, Phys. Plasmas 24, 072305 (2017). https://doi.org/10.1063/1.4989716
[5] C. Holland et al., Journal of Plasma Physics. 2023;89(4):905890418. https://doi.org/10.1017/S0022377823000843

Speaker: Eric Bass

82 Nonlinear wave-particle interaction and dynamics of energetic particle mode

Energetic particles (EPs) from fusion reactions and auxiliary heating can drive symmetry-breaking shear Alfvén wave (SAW) instabilities, including AEs and EPMs, via resonant wave-particle interactions, and understanding their nonlinear evolution is critical for understanding the alpha particle confinement in next-generation devices like ITER and CFEDR.
In our previous work [1], nonlinear simulations of Alfvénic instabilities in EAST ICRH experiments were carried out using the hybrid code MEGA [2]. The resonant excitation mechanisms of the dominant n=2 TAE-like EPM were elucidated, and nonlinear analysis revealed frequency bifurcation into high-frequency TAE and low-frequency BAE-like branches via upward and downward chirping, along with secular EP transport via phase-locking with the downward-chirping branch. However, the detailed mechanisms of the nonlinear frequency chirping remain to be fully understood.
To elucidate the underlying mechanisms, dedicated phase-space diagnostics [3] are employed in the present work to further investigate the nonlinear EPM evolution, based on the reference scenario studied in Ref. [1]. The results demonstrate that the nonlinear evolution is dominated not by a fixed cohort of initial resonant particles, but by the exchange of the resonant populations via self-consistent continuous trapping and de-trapping. This is evidenced by three representative wave-particle interaction behaviors: (i) initially resonant particles that become non-resonant and de-trapped; (ii) particles that remain trapped throughout; and (iii) initially non-resonant particles that become resonant during the nonlinear stage. In addition, the mode frequency chirping rate exhibits a linear scaling with the instantaneous mode amplitude, as predicted by the general theory [4,5]. Collectively, these results substantially advance the understanding of nonlinear EPM evolution [6], providing an intuitive physical picture fully consistent with the theory of nonadiabatic nonlinear evolution.
[1] Su P. et al 2026 Nucl. Fusion 66 022002
[2] Todo Y. et al 2015 Nucl. Fusion 55 073020
[3] Todo Y. et al 2021 Plasma Phys. Control. Fusion 63 075018
[4] Chen L. and Zonca F. 2016 Rev. Mod. Phys. 88 015008
[5] Zonca F. et al 2015 New J. Phys. 17 013052
[6] Su P. et al 2026 Nucl. Fusion (submitted)

Speaker: Dr Pengjuan Su (Key Laboratory of Frontier Physics in Controlled Nuclear Fusion and Institute of Plasma Physics, Chinese Academy of Sciences)

12:45 Lunch break

Poster session III and Coffee break

83 Destabilization of ctr-TAE modes: general analysis and a specific example

Mechanisms of destabilization of toroidal Alfvèn eigenmodes (TAE) in tokamaks are analyzed with the aim to reveal those leading to modes propagating in the direction opposite to plasma current, i.e. counter-propagating modes (ctr-TAE). Plasmas with fast-ions sources [such as neutral beam injection (NBI), ion cyclotron resonance heating (ICRH), fusion reactions] and without them (Ohmic discharges) are considered. A particular NSTX-U experiment with NBI, where co- and counter-propagating TAEs were observed simultaneously [1], is considered.
This presentation is based on the recent publication [2] which is extended to take into account finite width of the pitch distribution of the beam particles.
It was assumed in Ref. [2] that the pitch distribution of fast ions is approximated by the Dirac $\delta$-function, $F_\alpha\sim\delta(\chi-\chi_0)$, where $\chi=v_\|/v$, $v$ is the particle velocity, $v_\|$ is the velocity along the magnetic field. This approximation was justified by the fact that the dependence of the fast ion distribution on $\chi$ was not known. At the same time, it remained unclear whether finite width of distribution over $\chi$ does not stabilize the instability. In order to clarify this issue, the relations used in Ref. [2], which describe the instability growth rate, $\gamma(\chi)$, were extended by taking into account the presence of fast ions with various pitches, $\Delta\chi\neq0$. The numerical code calculating the growth rate was changed correspondingly. Calculations for both ctr-TAEs and co-TAEs were carried out for the cases when the fast ion density, $n_\alpha(r)$, has a plateau or a weak maximum, as shown in Fig. 4 of Ref. [2].
The results of calculations for two found ctr-TAE are shown in Fig. 1. We observe that finite $\Delta\chi$ decreases the growth rate, and can even completely suppress the instability (Fig. 1, left panel; when $\Delta\chi=0.2$, low mode is stable at $x_\max=0.35$ but unstable when the plateau is slightly extended, $x_\max=0.4$). On the other hand, it is a bit surprising that in some cases, when $\Delta\chi\neq0$, fast ions with pitches which are stabilizing for $F_\alpha\sim\delta(\chi-\chi_0)$ can become destabilizing (Fig. 1, left panel; the stable region of high mode at $\chi_0>0.75$ becomes unstable). This is explained by the fact that the positive derivative $\partial F_\alpha/\partial\chi>0$ shifts to smaller $\chi$ due to $\Delta\chi\neq0$.

Fig. 1. $\gamma$ versus $\chi_0$ for ctr-TAE in the NSTX-U discharge #205072 when $F_\alpha=n_\alpha(x)(v^3+v_c^3)^{-1}\exp[-(\chi-\chi_0)^2/\Delta\chi^2]$, with $x=r/a$: Left/top panel, $n_\alpha(x)$ with a plateau in the region $[0, x_\max]$, $x_\max=0.35$ except where noted $x_\max=0.4$. Right/bottom panel, $n_\alpha(x)$ with a maximum at $x_\max=0.35$.

[1] Podestà M., Fredrickson E.D. and Gorelenkova M. 2018 Nucl. Fusion 58 082023
[2] Kolesnichenko Ya.I., Fredrickson E.D., Lutsenko V.V., Tykhyy A.V. 2026 Nucl. Fusion 66 046002

Speaker: Anton Tykhyy (Institute for Nuclear Research, Kyiv, Ukraine)

84 Resonant mode-particle interaction priors for energetic particle phase-space tomography

A thorough understanding of the resonant interactions between fast
ions and plasma instabilities is vital for the successful operation of burning plasmas in future devices, such as ITER. It is well described that fast ions can drive instabilities, causing confinement degradation or, alternatively, turbulence suppression [1]. Measuring the fast-ion distribution function from experimental data is thus necessary for a complete understanding of the physics. However, this requires solving an ill-posed inverse problem. The resonant interactions between waves and fast ions, as well as fast-ion stability analyses, are conveniently described in terms of three invariants of motion, energy, toroidal canonical angular momentum, and magnetic moment, which renders three-dimensional reconstructions of the fast-ion distribution function highly beneficial. This in turn requires large quantities of data, of which we have a limited amount. Therefore, to facilitate the 3D reconstructions, sophisticated physics-informed priors must be developed to address the ill-posedness of the inverse problem. In this talk, we will describe recent work on 3D Bayesian inversions of fast-ion distribution functions
in constants-of-motion phase space from FIDA data on TCV, and we will compare the reconstruction quality using different priors. This includes a newly developed resonant mode-particle interaction prior based on the kick model [2–6], which builds upon recent work on incorporating radio-frequency wave heating as prior [7,8]. The use case adopted to test the methods is that of fast-ion-driven fishbone modes in the TCV tokamak.

[1] Salewski M. et al 2025 Nuclear Fusion 65 043002
[2] Podestà M., Gorelenkova M. and White R. B. 2014 Plasma Physics and Controlled Fusion 56 055003
[3] Podestà M., Gorelenkova M., Gorelenkov N. N. and White R. B. 2017 Plasma Physics and Controlled Fusion 59 095008
[4] Podestà M. et al 2019 Nuclear Fusion 59 106013
[5] Podestà M. et al 2021 Plasma Physics and Controlled Fusion 64 025002
[6] Podestà M. et al 2022 Nuclear Fusion 62 126047
[7] Rud M. et al 2025 Nuclear Fusion 65 056008
[8] Rud M. et al 2025 Nuclear Fusion 65 112006

Speaker: Mads Rud Larsen (École Polytechnique Fédérale de Lausanne)

85 Enhancing Tomographic Reconstructions of Energetic Particle Distributions Using Variational Auto Encoders

In future fusion power plants, to ensure safe and reliable operation, it will be necessary to be able to accurately and efficiently measure the distribution of energetic particles[1]. Since the fast-ion distribution cannot be measured directly, we need to reconstruct it from diagnostic measurements[2]. The problem of reconstructing the fast-ion distribution given measurements is an ill-conditioned inverse problem[3]. The set of diagnostic measurements used to reconstruct the fast-ion distribution is often sparse and noisy[4], leading to uncertain reconstructions that can be plagued by artifacts[5]. Improving both the quality and efficiency of such reconstructions is paramount to ensure a successful operation of future tokamaks such as ITER[6] and SPARC[7]. In this work, we discuss a possible solution in the form of variational auto encoders (VAE). VAEs are a type of deep neural network that have been used to solve problems such as de-noising[8], feature extraction[9] and image classification. We show that VAEs can be used to post-process tomographic reconstructions of the fast-ion distribution to achieve identification and removal of artifacts. We also show that VAEs can be used to reduce the dependence of tomographic reconstructions on the choice of the regularization parameter $\lambda$.

[1] M. Salewski, Fast-ion diagnostic in fusion plasmas by velocity-space tomography, Technical University of Denmark, 2019
[2] M. Salewski, et al, 2019, JINST 14 C05019
[3] M. Rud et al, 2026, Plasma Phys. Control. Fusion 68 015035
[4] B. Madsen et al, 2020, Plasma Phys. Control. Fusion 62 115019
[5] H. Järleblad et al, 2025, Nucl. Fusion 65 016060
[6] B. Bigot et al, 2019, Nucl. Fusion 59 112001
[7] A.J. Creely et al, 2020, Plasma Phys. Control. Fusion 86 865860502
[8] S. Ruipérez-Campillo et al, 2025, Expert Systems With Applications 300 130185
[9] H. Wang et al, 2025, Energy 330 136716

Speaker: Dr Henrik Järleblad (Department of Physics, Technical University of Denmark)

86 Alpha Particles without DT: Alternative Routes in Future JT-60SA Plasmas

Speaker: Rui Coelho (Instituto Superior Tecnico - IPFN)

87 Experimental characterization of fishbones and associated energetic particle transport via correlated FIDA and FILD measurements on the TCV tokamak

Speaker: Mr Quentin Kernel (EPFL - Swiss Plasma Center)

88 Measurements of Fast Ion Distribution During the Alfvén Eigenmode-Induced Microturbulence Suppression in DIII-D

Speaker: Xiaodi Du (GAT - Fusion Energy Research General Atomics (US))

89 Phase-space mechanisms of ICRF-driven bursting shear Alfvén waves in tokamak plasmas

Speaker: Jialei Wang (National Institute for Fusion Science)

90 Properties of Alfvén Eigenmodes during Strong Microturbulence Suppression

Speaker: William Heidbrink

91 Qualitative validation of CAD-based simulations of NBI-induced prompt fast-ion losses against IRTV measurements in KSTAR

Speaker: Taeuk Moon (Korea Institute of Fusion Energy)

92 Simulation study of nonlinear wave-particle interaction of energetic particle modes in tokamak plasmas

Speaker: Pengjuan Su

93 TGLF-EP+Alpha reduced fast ion transport prediction validation against diffusive transport measurements

Speaker: Kyle Callahan (UCI)

94 Nonlinear frequency chirping of energetic-particle-driven Alfvénic fluctuations in ORB5 gyrokinetic simulations

Frequency chirping of Alfvénic modes is a ubiquitous signature of the nonlinear phase-space dynamics of energetic particles (EPs) in tokamaks [1], and the chirping rate carries direct information on the resonant EP redistribution that limits fast-ion confinement [1]. Recent work shows that the nonlinear dynamics of the thermal plasma — beat-driven zonal fields and zonal currents together with the phase-space zonal structures of the thermal species — plays an essential role in the chirping and saturation of Alfvén eigenmodes, in some regimes dominating over the EP nonlinearity [2, 3]. We present a systematic global gyrokinetic study of chirping toroidal Alfvén eigenmodes (TAEs) with the particle-in-cell code ORB5, addressing how the chirping rate is set by the EP drive and how it is modified when several toroidal mode numbers are allowed to interact.
The simulations use a circular, large-aspect-ratio equilibrium with a monotonic safety-factor profile, a flat bulk-plasma density and a core-peaked energetic-ion population whose concentration is scanned to vary the linear drive. A scan over toroidal mode number identifies a most unstable intermediate-n TAE. The saturation amplitude nevertheless decreases monotonically with toroidal mode number, indicating that the radial width of the mode envelope, rather than the linear growth rate, sets the attainable mode level.
In the nonlinear phase all cases chirp downward, the sense expected both from the evolution of resonant EP phase-space structures [1] and from zonal-current-mediated frequency shifts [2]. Tracking the instantaneous frequency [4], the chirping rate increases with the EP drive, but distinctly more slowly than the adiabatic phase-space-structure scaling near marginal stability [1, 4]; the saturation level likewise grows only weakly with the linear growth rate. The central result is that, within each simulation family, the chirping rate is a linear function of the saturated mode amplitude, consistent with the chirp being governed by the wave-trapping frequency of the resonant EPs rather than by the linear drive itself [1, 4].
Multi-mode simulations show that this coupling is strong. Retaining the axisymmetric zonal component alone raises both the saturation level and the chirping rate of the driven TAE severalfold relative to the single-n case at the same drive, in line with the zonal-field-mediated enhancement of TAE saturation found when thermal-plasma nonlinearities are retained [3]. Adding further toroidal harmonics then reduces both quantities monotonically, as the modes compete for the same EP free energy, while all cases remain on a common chirp-rate–amplitude line. Single-n predictions of chirping rates, and hence of the inferred EP transport, can therefore be substantially in error: zonal flows, zonal currents, mode–mode coupling and the nonlinear thermal-plasma response [2, 3] must be retained in quantitative modelling of chirping AEs [1].
References
[1] L. Chen and F. Zonca, “Physics of Alfvén waves and energetic particles in burning plasmas”, Rev. Mod. Phys. 88, 015008 (2016).
[2] R. Ma, P. Liu, L. Chen, F. Zonca and Z. Qiu, “How zonal fields suppress reversed shear Alfvén eigenmode in tokamak plasmas”, Phys. Rev. Lett. 137, 045101 (2026).
[3] N. Chen, T. Hayward-Schneider, F. Zonca et al., “On nonlinear saturation of toroidal Alfvén eigenmode due to thermal plasma nonlinearities”, arXiv:2604.15024 (2026).
[4] R. Wu, A. Biancalani, M. V. Falessi et al., “Frequency chirping of energetic-particle-driven geodesic acoustic modes in tokamaks”, arXiv:2603.23430 (2026).

Speaker: Xin Wang

95 Nonlinear saturation of toroidal Alfven eigenmode by self-organized zonal fields

Recent fusion experiments have observed significant improvement in plasma confinement in the presence of fusion-born alpha particles [1,2], with Alfven eigenmodes playing a central role in regulating drift-wave microturbulence. First-principles gyrokinetic theory has proposed two mechanisms: direct three-wave interaction [3] and indirect interaction mediated via zonal fields or phase-space zonal structures (PSZS) [4], with the latter attracting increasing attention in both experimental and simulation studies.

In this work, we employ the global gyrokinetic code GTC to investigate the beat-driven excitation of zonal fields by toroidal Alfven eigenmodes (TAEs). In particular, the zonal current (ZC), characterized by radially fine-scale structures, shows good agreement between simulation and theoretical predictions [5]. This ZC further induces local modification of the q-profile within the TAE saturation amplitude (delta_B/B_0~10^-3), and may potentially break the radial correlation of microturbulence and reduce bulk plasma transport through oscillating magnetic shear. The zonal flow (ZF) exhibits both meso-scale and fine-scale structures, attributed to contributions from energetic particles (EPs) and thermal ions. The ZF plays a dominant role in TAE saturation by influencing bulk plasmas, while also playing a subdominant role that weakly increases the TAE linear growth rate via its effect on EPs [5]. Progress in the verification and validation of TAE-driven zonal fields through gyrokinetic theory and simulation will be reported.

Reference
[1] Mazzi S et al. 2022 Enhanced performance in fusion plasmas through turbulence suppression by megaelectronvolt ions Nature Physics 18 776–782
[2] Ruiz Ruiz J et al. 2025 Measurement of zero-frequency fluctuations generated by coupling between alfven modes in the jet tokamak Phys. Rev. Lett. 134 095103
[3] Chen L et al. 2023 On nonlinear scattering of drift wave by toroidal Alfven eigenmode in tokamak plasmas Nucl. Fusion 63 106016
[4] Fang Q et al. 2025 Indirect nonlinear interaction between toroidal Alfven eigenmode and ion temperature gradient mode mediated by zonal structures Nucl. Fusion 65 066004
[5] Chen L et al. 2025 The effects of zonal fields on energetic-particle excitations of reversed-shear Alfven eigenmode: simulation and theory Nucl. Fusion 65 016018

Speaker: Dr Jian Bao (Institute of Physics, Chinese Academy of Sciences)

96 A two-fluid model of the energetic-particle-driven geodesic acoustic mode

We have developed a two-fluid model of the energetic-particle-driven geodesic acoustic mode (EGAM) in magnetically confined toroidal plasmas by applying fluid models to the bulk plasma and energetic particles separately. Analysis of the theoretical model reveals two EGAM solutions with high and low frequencies. In the high-frequency mode, the pressure perturbations of the two components oscillate in phase, whereas they oscillate out of phase in the low-frequency mode. In this presentation, the analytical results are compared with gyrokinetic theory [1] and simulation results [2].

We consider an electrostatic magnetohydrodynamic (MHD) oscillation in a toroidal plasma with concentric circular magnetic surfaces. Let $r$ denote the minor radius, $q(r)$ the safety-factor profile, and $R_0$ the major radius at the plasma center. The bulk plasma and energetic particles are denoted by the subscripts $b$ and $h$, respectively. The equilibrium density and pressure profiles are denoted by $\rho_{s0}(r)$ and $P_{s0}(r)$ ($s=b$ or $h$), and a static equilibrium with no plasma flow is considered.

By coupling the linear equations for the radial electric field, the pressure perturbations of the two components, and their velocity perturbations parallel to the magnetic field, we analyze an axisymmetric mode (toroidal mode number $n=0$) under the large-aspect-ratio approximation ($R_0 \gg r$). When the poloidal mode number of the radial electric field is $m=0$, the pressure and parallel-velocity perturbations depend on $\sin \theta$ and $\cos \theta$, respectively. The analysis yields the following solutions for the eigenfrequency $\omega$.
\begin{equation}
\omega^2 = \frac{1}{2}\left( \omega_{Gb}^2 + \omega_{Gh}^2
\pm \sqrt{( \omega_{Gb}^2 - \omega_{Gh}^2)^2 + 4\Omega_b^2 \Omega_h^2} \right) \ .
\end{equation}
Here, $\omega_s^2 \equiv \gamma_s P_{s0}/(\rho_{s0} q^2 R_0^2)$, $\Omega_s^2 \equiv 2 \gamma_s P_{s0}/(\rho_{0}R_0^2)$, and $\omega_{Gs}^2\equiv \omega_s^2 + \Omega_s^2$.
Note that $\Omega_s$ is defined using the total density $\rho_0 \equiv \rho_{b0}+\rho_{h0}$ rather than the density $\rho_s$ of each component. Here, $\gamma_s$ is the ratio of specific heats. In the limit where the energetic-particle density, pressure, and $\Omega_h$ are negligibly small ($\rho_{h0}\simeq 0$, $P_{h0}\simeq 0$, $\Omega_{h}\simeq 0$), the two solutions converge to
\begin{equation}
\omega = \omega_\mathrm{GAM}\equiv \frac{1}{R_0} \sqrt{ \frac{\gamma_b P_{b0}}{\rho_{b0}} \left(2+\frac{1}{q^2} \right)} \ , \qquad
\omega = \omega_h \ .
\end{equation}
The first solution is the geodesic acoustic mode (GAM) in the absence of energetic particles. The second solution is the frequency of the energetic-particle acoustic wave, which arises from compression associated with the parallel flow velocity.

[1] G.Y. Fu, Phys. Rev. Lett. 101, 185002 (2008).
[2] Hao Wang et al., Phys. Rev. Lett. 120, 175001 (2018).

Speaker: Prof. Yasushi Todo (National Institute for Fusion Science)

97 Development of scenarios and diagnostics for fast ion studies in next generation tokamaks without beryllium

Following the end of the operations of the Joint European Torus (JET) and the recent development of a new baseline for ITER, future tokamaks will be based on a high Z first wall material and will no longer use beryllium. New developments in fast ion generation scenarios and neutron and gamma-ray diagnostics are necessary to provide both the instruments and the knowledge base required in preparation of experiments at these devices. In this contribution, we present the first numerical and experimental results of a mid-term program aimed at advancing fast-ion scenarios for α particle generation and the related diagnostics, mostly in view of JT-60SA and ITER operations.
On the experimental side, scenarios for α particle production based on the 3He(d,p)α or the 11B(p,α)2α reactions are being designed and first exploratory tests have been made at the ASDEX Upgrade tokamak. To this end, we have developed a new compact neutron spectrometer, named COSMONAUT+, based on a LaCl3(Ce) scintillator. The instrument has successfully measured, for the first time at ASDEX Upgrade, the neutron spectrum from the acceleration of fast deuterium ions by third harmonic radio-frequency heating. This is a necessary step to demonstrate the generation of α particles from D+3He fusion in forthcoming experiments in D-3He plasmas. A gamma-ray spectrometer has also been installed to enable gamma-ray emission measurements from fast-ion reactions without beryllium in next step experiments.
On the numerical side, we have evaluated the velocity-space and orbit-space sensitivity of boron-based gamma-ray spectroscopy using the 10B(α, pγ)13C reaction, which is one of the options for α particle measurements in tungsten first wall devices with boron injection. Despite a lower signal when compared with the 9Be(α, nγ)12C reaction used at JET, the three emission peaks of 10B(α, pγ)13C offer a selective sensitivity to super- and sub-Alfvénic velocities, with a contribution coming mostly from particles on trapped and potato orbits, for the case of radial lines of sight. This is a unique feature of 10B(α, pγ)13C, which was not available by 9Be(α, nγ)12C, and may be exploited for α particle studies in low neutron yield scenarios and in preparation of high-power deuterium-tritium experiments.

Speaker: Massimo Nocente (Instituto Nazionale di Fisica Nucleare (EU))

98 Effects of NBI generated energetic particle distribution on reversed shear Alfvén eigenmodes

The kinetic-magnetohydrodynamic hybrid code, MEGA, is employed to study the effect of energetic particle (EP) distribution on reversed shear Alfvén eigenmodes (RSAEs). Focusing on the most unstable (n=2) RSAE in an EAST discharge, the EP distribution is represented as a linear sum of multiple Gaussian functions, following Gorelenkov [Nucl. Fusion 45, 226 (2005)] and Liu [Phys. Plasmas 21, 056105 (2014)]. This modification concentrates more EPs along the NBI direction, consistent with realistic observation in the experiment, and yields the phase-space distribution closer to NUBEAM output, thus better explaining the enhanced RSAE instability when the beam switches from tangential to perpendicular injection. The linear growth rate is larger with co-injection than counter-injection, which is especially sensitive to the peak of the EP distribution in pitch angle space close to the dominant resonance band. For this reason, the saturation mechanism is radial decoupling for co-injection. The frequency and amplitude have drastic fluctuations since mode structure varies significantly. For counter-injection, the saturation mechanism is resonance detuning due to low linear growth rate so that the mode structure largely preserves its shape. The energy transfer from each component is comparatively lower, therefore does not induce significant amplitude fluctuations. In addition, counter-injection case has higher saturated level of zonal current, related to the RSAE evolution, and more EPs are confined due to the weaker interaction between mode and particle.

Speaker: Yunpeng Zou (Institute of Plasma Physics, Chinese Academy of Science)

99 Electron Cyclotron Heating Modification of Toroidicity Induced Alfvén Eigenmode Activity in DIII-D

Localized electron cyclotron heating (ECH) can have a dramatic effect on neutral beam driven toroidicity induced Alfvén eigenmode (TAE) activity in DIII-D plasmas. In these experiments, ECH with varied current drive phasing was deposited at large radius near the location of TAEs that have been recently reported to locally stabilize ITG turbulence through the generation of sheared flows [1]. The impact of ECH depends drastically on location, with an increase in mode activity observed inside of the deposition location while a local reduction in some TAEs occurs near the deposition location. Modes that are stabilized return rapidly when electron cyclotron power is removed and the effect does not depend on current drive phasing, which indicates the minimal current driven at large radius does not play a large role. In some cases turbulence suppression persists through the EC injection period. In other cases, with EC injection inside of the TAE generated ITB, the electron temperature rises rapidly until the ITB collapses and both the TAEs and turbulence suppression cease. The rapid change in TAE activity with both EC turn-on and turn-off indicate electron collisional damping or electron Landau damping may play a role as opposed to current profile modification effects. To this end, NOVA-K calculations of TAE stability have been carried out for a range of conditions approximating those during a radial and power scan of ECH deposition in the target DIII-D equilibrium. Electron collisional and Landau damping are two of the dominant damping mechanisms found and a factor of three or more variation is predicted to be possible depending on deposition location and power.

*This work was supported by the US Department of Energy under DE-FC02-04ER54698 and DE-SC0026408 and DE-AC02-09CH11466

[1] X.D. Du, et al., Phys. Rev. Lett.66 135, 265101 (2025).

Speaker: Michael Van Zeeland (GAT - Fusion Energy Research General Atomics (US))

100 Energetic-ion Driven Instabilities during I-phase and inter-ELM in the HL-3 Tokamak

In HL-3 H-mode plasma discharges, two new MHD instabilities have observed and identified. Both instabilities are localized in the outer region or the pedestal region of the plasma. The first is a low-frequency instability with f=0–50 kHz, exhibiting rapid frequency chirping and a toroidal mode number of n=3. It propagates in the ion diamagnetic drift direction, displays energetic-particle mode (EPM) behavior, and triggers I-phase oscillations. The second is a high-frequency instability with f=100–500 kHz. The most unstable mode has a toroidal mode number of n=0. This instability family sometimes manifests as three frequency bands exhibiting an integer harmonic relationship. The most unstable mode frequently couples with an n=1 BAE mode in the outer region to generate numerous sideband modes. These instabilities tend to appear during the ELM-free phase and preceding ELMs, but can also be observed in Ohmic heating plasma, albeit with weaker amplitude. This most unstable mode exhibits global Alfvén eigenmode (GAE) characteristics.

Speaker: Dr Xiao-Long Zhu (Southwestern Institute of Physics)

101 Global and local monitoring of fast-ion losses in the presence of external magnetic perturbations in the MAST-U spherical tokamak using IR thermography and a Fast Ion Loss Detector

Instabilities such as Edge Localised Modes (ELMs) can degrade confinement in tokamak plasmas [1, 2]. Externally applied Magnetic Perturbations (MPs) are used to suppress these instabilities [3]. This technique is still under investigation in many tokamaks, and its understanding is crucial for future fusion devices, as applying these perturbations can also have a detrimental effect on fast-ion confinement [4]. The MAST-U spherical tokamak is equipped with an upper and lower sets of MP coils, and fast-ion losses are directly measured and analysed in this device using a scintillator-based Fast-Ion Loss Detector (FILD) [5]. FILD measurements reveal a clear modulation of the total losses with the applied configuration of the MPs (perturbation amplitude, poloidal spectrum, and phase) and with the velocity-space of the fast-ion losses [6].
However, FILD measurements only reflect local behaviour of the losses. Infrared (IR) thermography can be used to complement FILD local measurements [7], as it monitors a wider region of the tokamak wall, allowing a more global understanding of the behaviour and deposition pattern of the fast-ion losses [8, 9]. To this end, a temporal modulation of the NBI source that provides the fast-ion distribution has been employed. The IR measurements enable the qualitative identification of the regions where the fast-ion losses are deposited. The application of this method indicates that the local FILD measurements should not be treated as indicative of the MP configuration that optimises fast-ion confinement: The IR data shows that, under several configurations, the applied perturbation leads to lost fast-ions impinging on different regions of the wall, and not necessarily to a global improvement or deterioration of the confinement. The experimental results are compared with calculations performed using MARS-F to model the plasma response to the 3D perturbation and ASCOT to reproduce the changes in the global deposition patterns of the fast-ion losses.

[1] M. Garcia-Muñoz et al 2013 PPCF 55 124014
[2] J. Coenen et al 2015 JNM 463 7884
[3] T.E. Evans et al 2004 PRL 92(23) 235003
[4] L. Sanchis et al 2021 NF 61 046006
[5] J. F. Rivero-Rodríguez et al 2018 RSI 89 10I112
[6] L. Velarde et al 2025 NF 65 112003
[7] L. Velarde et al 2025 FED 220 11
[8] C.J. Lasnier et al 2014 RSI 85 11D855

Speaker: Lina Velarde (Universidad de Sevilla)

102 Hybrid Kinetic-MHD Analysis of Energetic Particle (1,1) Modes in SPARC

Linear (n=1) hybrid kinetic-MHD simulations have been performed with the NIMROD code to investigate energetic particle effects on (m=1,n=1) stability in SPARC plasmas. Understanding energetic particle-driven instabilities is critical for burning tokamaks, where fusion-born alphas can strongly modify the stability of MHD modes, impacting plasma confinement and performance. Using the SPARC Primary Reference Discharge
[ https://github.com/cfs-energy/SPARCPublic/tree/main/PrimaryReferenceDischarge ]
($B_0$=12T, I=8.7MA, $q_0$=0.93, $\beta_0$=4.5$\%$), a scan of energetic particle $\beta_\mathrm{hot}$ is carried out and growth rates and real frequencies are measured.

These simulations demonstrate that increasing $\beta_\mathrm{hot}$ stabilizes the ideal internal kink mode while simultaneously destabilizing an energetic particle-driven fishbone mode, illustrating the typical (1,1) transition from an MHD-dominated instability to a kinetic energetic particle mode [C. C. Kim, PoP 15 (2008)]. The NIMROD results are compared with corresponding linear calculations performed using the M3D-C1 hybrid kinetic-MHD model [C. Liu, CPC 275 (2022)], providing a cross-code benchmark of the growth rates and mode frequencies; additional validation of US flagship hybrid kinetic-MHD modeling tools for predicting energetic particle modes in next-generation fusion devices such as SPARC and ARC.

Phase-space diagnostics are presented to identify the dominant particle resonances that interact with the kink/fishbone mode. Both trapped and passing particles are shown to play a role in the evolution of the (1,1) mode. With these diagnostics, we highlight the localization and intrinsic connection of the resonant energetic particle orbits in both physical and velocity space. This multi-dimensional (5D) analysis provides an expanded view into the drift kinetic orbit topologies of energetic particles and theirs interaction with the (1,1) mode and creates some foundation for a bridge between the abstracted and idealized analysis of pencil-and-paper theory (e.g. zero width bananas) and the complications of real world experiments and their diagnostics.

Speaker: Charlson Kim (SLS2 Consulting)

103 Phase- and velocity-space sensitivity of collective Thomson scattering measurements in magnetized plasmas

Collective Thomson scattering (CTS) diagnostics have been installed at several machines and will be primary alpha-particle diagnostic at ITER. CTS measurements are sensitive to the projection of the local velocity distribution function in a measurement volume. The thermal ions with low projected velocities dominate the low-Doppler-shift part of CTS spectra whereas the fast ions with high projected velocities dominate the high-Doppler-shift parts. The projection determines to which extent a given region in fast-ion phase- and velocity-space contributes to a given spectral bin of a CTS spectrum. Here, we discuss the formation of the high-Doppler-shift parts of CTS spectra as well as the corresponding weight functions that quantify the velocity- and phase-space sensitivities. The weight functions can be split into a rate part that depends only on the frequency or, equivalently, the projected velocity, and a projection part from either 2D velocity space or 3D phase space. CTS weight functions allow rapid interpretation of CTS spectra and the inversion of CTS spectra in velocity space or phase space in conjunction with other energetic particle diagnostics.

Speaker: Mirko Salewski (Technical University of Denmark (ZZ))

104 Plasma shaping effects on fast particles and Alfvén eigenmodes dynamics in spherical tokamaks

Spherical tokamaks (STs) represent one of the most promising research pathways towards fusion energy. Owing to their low aspect ratio, devices such as the Mega Amp Spherical Tokamak Upgrade (MAST-U) allow access to high-β plasma regimes, potentially offering a route towards more compact and efficient fusion reactors.
In spherical tokamaks such as MAST-U, Alfvén Eigenmodes (AEs) can play an important role, as the ST configuration strongly influences the Alfvén continuum, favouring the excitation by fast ions of global coherent modes, such as Toroidal Alfvén Eigenmodes (TAEs), for which the toroidicity-induced frequency gap is enhanced.
Despite AEs being widely recognised as an important factor in the confinement of both energetic particles and bulk plasma, in spherical tokamaks their role remains only partially explored compared to large-aspect-ratio devices. In particular, in spherical tokamaks such as MAST-U, the combination of low aspect ratio, strong natural plasma shaping, and high-β values introduces a regime in which AEs–energetic-particles interactions are expected to be strongly dependent on plasma shaping parameters, such as elongation and triangularity, which modify the Alfvén continuum, poloidal mode coupling, and the spatial structure of the eigenmodes.
The quantitative impact of plasma shaping on Alfvén Eigenmode stability around MAST-U like configurations has been investigated in both the linear and nonlinear regimes using the ORB5 gyrokinetic code, with particular emphasis on dynamical phenomena such as frequency chirping and convective losses of energetic ions. These aspects are essential for the development of reliable predictive models of wave–particle interaction and for assessing the robustness of the spherical tokamak concept in reactor-relevant burning plasma regimes.

Speaker: Francesco Palermo (UKAEA)

105 The Charged Fusion Product Detector (CFPD) on MAST-U

The Charged Fusion Product Detector (CFPD) is a fast-ion diagnostic employing solid-state detectors to measure energies and production rates of charged fusion products [in particular p (3.02 MeV) and t (1.01 MeV) from D-D fusion] primarily from fast-ion induced reactions. Results obtained during the recent MU04 and MU05 (2024-26) MAST-U campaigns will be presented. The current diagnostic is an advancement from a previous version [1]: with improved signal-to-noise ratio reducing the spectroscopic energy-peak width and allowing lower energy products to be measured.

Radial distributions of fusion-product emissivities measured using a 6-channel, all diamond-detector set-up will be shown. Also, fast-ion populations have been studied using silicon-detectors via spectroscopy of the charged fusion product energy peaks, exploiting line-of-sight motions of reacting fast-ions that affect the energy distribution of the products. Results will be compared with TRANSP/FIDASIM predictions [2,3].

[1] A. Aboutaleb et al. Rev. Sci. Instrum. 95, 083522 (2024)
[2] B. Geiger et al. PPCF 62, 105008 (2020)
[3] W. W. Heidbrink et al. PPCF 63, 055008 (2021)

Speaker: Edward Parr (UKAEA, Abingdon, Oxfordshire, UK)

EP in stellarators and 3D configurations

106 Finite Larmor radius effects in stellarators – the hidden ripple inside hidden symmetry

It has long been understood that alpha particle equilibrium in tokamaks relies on the existence of approximate symmetries in time, gyroangle and toroidal angle. Violation of these symmetries in tokamaks leads to energetic particle losses observed in simulation and experiment. This behaviour is well modelled in terms of collisionless particle trajectories.
In stellarators, a chief concern for confinement is to adequately substitute for these symmetries when nominal axisymmetry is not available. A number of schemes have been proposed, but underlying them all is the realization that the modulus of B is the most important determinant of the guiding centre motion, and the quantity upon which a symmetry must be constructed. Quasi-symmetric solutions to the force balance equation in stellarators have been found by Landreman and co-authors which promise integrable guiding-centre motion. A key question is whether the full-orbit motion remains in the neighbourhood of the perfectly confined guiding-centre orbits or whether a secular component to the trajectory accumulates to deconfine the particles. Expensive Monte-Carlo simulations can always give a definitive answer to this question for any given case under scrutiny, but insights into the reasons can be elusive, and the optimizations required for improvement are not obvious.
Exploiting the similarity between quasi-axisymmetric motion and tokamak motion, we have constructed a novel method to capture the full orbit corrections as a fictitious magnetic perturbation. All deviations due to ripples, error fields, Alfven waves, and FLR effects may therefore be treated in a unified manner. The most important component of this fictitious field is revealed to be one that does not depend on gyroangle and relates to the torsion of the magnetic field 𝜏≡(𝑏̂⋅∇𝑒̂2)⋅𝑒̂1. Accumulating guiding centre trajectory error can be expected for resonant orbits in the usual manner once drift-island overlap occurs. The magnitude of the fictious ripple is approximately 𝛿𝐵𝐵∼𝜏𝐿𝜌𝛼∗ 2 suggesting a crude estimate of 𝜌𝛼∗ ∼10% for island overlap resulting in alpha loss. Details of the torsion and how the resonance width depends on shear are likely to lower this estimate in specific cases.
This work has been funded by the EPSRC Fusion Grant 2022/27 [grant number EP/W006839/1]. To obtain further information on the data and models underlying this paper please contact PublicationsManager@ukaea.uk*.

Speaker: Michael Fitzgerald (UKAEA)

107 Energetic particle transport induced during core density collapses in LHD and Stellarator reactor relevant configurations

Core density collapses (CDC) events observed in the Large Helical Device (LHD) are a major concern for the performance of future Stellarator reactors, particularly for quasi-symmetric magnetic field configurations with weak magnetic shear. CDC severely restrict the maximum β of LHD outward shifted magnetic axis configurations, strongly fueled by pellets and heated by neutral beam injector (NBI), leading to a peaked pressure profile and unstable ballooning modes at the plasma edge [1,2]. Nonlinear FAR3d simulations, based on previous CDC modeling studies [3], are performed to analyze the transport of energetic particles (EP) during CDC events. Three different EP populations are considered in the analysis: Passing EP injected by the tangential NBI, helically trapped EP injected by the perpendicular NBI as well as a theoretical reactor relevant case with alpha particles. Passing EP simulations show 50% losses induced by the combination of saturating middle n ballooning modes in the middle plasma region and low n modes at the inner plasma, leading to ballistic EP transport linked to the stochastization of the magnetic field. CDC precursors and low n perturbation during the CDC induce intense EP losses and fluxes. Trapped EP simulation indicates an almost depletion of the EP population, 80% losses, caused by the magnetic field stochastization at the plasma edge during the ballooning modes saturation. CDC precursors induce ballistic transport and large trapped EPs fluxes before the CDC destabilization. FIDA measurements indicates a decrease of the EP content at the plasma core, mainly passing EPs generated by the tangential NBI, once the CDC is triggered. The radiance is reduced to the half after the CDC, consistent with the passing EP losses obtained in the simulation. Consequently, both CDC and precursors, reduce the device plasma confinement and heating performance. The simulation of the theoretical model with alpha particles shows 50% losses, induced by the combination of saturating middle n ballooning modes in the middle plasma region and low n modes at the inner plasma, leading to a ballistic alpha particle transport linked to the stochastization of the magnetic field, similar to the passing EP case. Nevertheless, the largest losses and fluxes are induced by the CDC precursors. That means, CDC can severely deteriorate the performance of future Stellarator reactor limiting their economical viability and must be avoided.

[1] S. Ohdachi et al, Contrib. Plasma Phys. 50, 552 (2010).
[2] S. Ohdachi et al, Nucl. Fusion 57 066042 (2017).
[3] J. Varela et al, Nucl. Fusion, 66, 026017 (2026).

Speaker: Jacobo Varela Rodriguez (Institute for Fusion Studies UT-austin)

108 Suppression of Coherent Energetic Particle Phase Space Structures due to 3D fields

Modulations in amplitude and phase of Alfvénic eigenmodes (AEs) may occur due to wave particle interactions with energetic particle (EP) distribution functions [1]. Damping can shrink the resonance with the mode, creating a hole-clump pair in the region of the shrunken resonance [1]. Hole-clump pairs in phase space are associated with phase modulations (chirps) of modes and propagate through phase space, increasing transport [1], and have experimentally observed in chirping fishbone modes to expel EPs to the wall in phase with the mode [2]. Previous work has proved that stochastically diffusive behaviors, such as microturbulence, can suppress chirping modes, most likely by diffusing the relevant hole-clump pair [3]. Additionally, 3D error fields in NSTX have been shown to suppress chirping in AEs, inducing mode saturation, suggesting the use of 3D fields for ‘phase space engineering,’ to intentionally suppress targeted modes [4]. Imperfections in omnigenous fields in stellarators, and 3D effects more generally, can result in new classes of trapped particles with unique convective drifts [5]. These drifts can introduce chaos around resonances, introducing a stochasticity that can sometimes act as an effective diffusion. In this work, we study the diffusive effects of error in omnigenous fields and 3D fields on energetic particles, we measure phase space stochasticity using Weighted Birkhoff Averaging (WBA), introduce a numerically computed effective diffusion coefficient, and compare to the stochastic effects of pitch angle scattering.

[1] Bierwage,A., et al. Plasma & Fus. Res. 16,(2021):1403087.
[2] Heidbrink, W., et al. Plasma Phys. Control. Fusion. 53, (2011) 085028.
[3] Duarte,V., et al. PoP 24, 122508(2017).
[4] Bortolon, et al. PRL. 110, 265008 (2013).
[5] Paul,E.J., et al. NF. 62, n.12(2022):126054.

Speaker: Alexandra Lachmann (Columbia University)

19:00 Conference dinner

Friday 2 October

Fast electrons and runaways

109 The Effect of Externally Applied and Self-Excited Waves on Relativistic Electrons

Wave–particle interactions have emerged as a promising pathway for directly affecting runaway electron (RE) dynamics and enabling new mitigation strategies in tokamaks. Recent experiments on the DIII-D tokamak explored both externally launched and self-generated waves that resonantly interact with the RE population.

Experiments utilizing externally launched electron cyclotron (EC) waves investigated a novel mitigation scenario based on OXB mode conversion. When O-mode waves were launched at the optimal angle relative to the magnetic field, a strong density enhancement formed near the predicted conversion window, indicating successful generation of the slow-X mode. Application of EC power during the RE plateau modified the loop voltage and synchrotron emission, consistent with possible changes to RE dynamics.

More recent experiments focused on externally launched helicon waves, which were observed to drive significant pitch-angle scattering of runaway electrons in the low-density Quiescent Runaway Electron Regime (QRE). These interactions produced a momentum-space vortex structure that imposed a new upper limit on RE energies, in agreement with kinetic modeling and theoretical predictions. The results provide direct experimental evidence that helicon waves can strongly modify relativistic electron phase-space evolution.

In addition to externally launched waves, self-excited runaway-driven whistler waves were observed to resonantly interact back on the RE population. Fast visible camera measurements revealed a predator–prey-like relationship between the wave activity and RE synchrotron emission, where wave growth leads to enhanced pitch-angle scattering and synchrotron damping of the RE beam, followed by re-acceleration and renewed wave growth. These observations provide insight into nonlinear wave–particle feedback mechanisms that may play an important role in regulating RE dynamics.

Together, these experiments advance the understanding of fundamental wave–particle interactions in RE plasmas and demonstrate their potential relevance for disruption mitigation in future devices.

Work supported by US DOE under DE-FC02-04ER54698, DE-SC0021622, DE-FG02-07ER54917, DE-AC52-07NA27344, and DE-SC0022270

Speaker: Alexander Battey (Ecole Polytechnique Federale de Lausanne (EPFL), Swiss Plasma Center (SPC), CH-1015 Lausanne, Switzerland)

110 Crested stationary spectra of weak turbulence

Energetic-particle-driven instabilities in plasmas commonly modify the particle distribution function to establish a balance between the instability drive and the wave-damping rate. This balance can last long in the presence of particle sources and sinks. In such situations, a stationary spectrum of turbulence and the distorted distribution of particles need to be determined self-consistently. The idealized one-dimensional descriptions of such systems do not generally apply to two- or three-dimensional systems of practical interest, which calls for more relevant theoretical modeling. This talk shows that the stationary spectra of turbulence in such systems do not occupy two- or three-dimensional areas in phase space but reside on lines or surfaces and look like a mountain crest. A recent example is in Ref. [1], presenting a quasilinear theory for stationary whistler turbulence driven by runaway electrons. Such stationary spectra also form when stimulated scattering on plasma ions balances the excitation of plasma waves by a relativistic elec-tron beam [2]. Technically, the crested nature of the wave spectrum reduces the dimensionality of the problem and facilitates a self-consistent description of the turbulence and its impact on the driving elec-trons, both analytically [1, 2] and numerically [3].

[1] B.N. Breizman and D.I. Kiramov, Marginal stability constraint on runaway electron distribution, Physics of Plasmas 30, 022301(2023).
[2] B.N. Breizman, V.E Zakharov, and S.L. Musher, Kinetics of Stimulated Scattering of Langmuir Waves by Plasma Ions, Soviet Physics JETP 37, 658 (1973).
[3] K. Huang and B. N. Breizman, The marginally stable state of runaway electrons controlled by self-excited waves, Poster presentation 1P060 at the 51st EPS Conference on Plasma Physics (Vilnius, Lithuania, July 7-11, 2025).

Speaker: Boris Breizman (The University of Texas at Austin)

111 Nonlinear coupling between beta-induced Alfvén eigenmodes and toroidal Alfvén eigenmodes in the EAST tokamak

Multiple low-frequency beta-induced Alfvén eigenmodes (BAEs, 10 - 20 kHz) and a high-frequency toroidal Alfvén eigenmode (TAE, 120 - 145 kHz) driven by runaway electrons are simultaneously observed in the EAST low-density Ohmic discharges. A new high-frequency mode (HFM, 100 - 170 kHz), generated by nonlinear coupling between BAEs and TAE, are conclusively identified.
The experiments were performed on the EAST tokamak in ohmic plasmas. As the electron density decreases to approximately 0.41019 m–3, a large population of REs are generated. Multiple low-frequency BAEs and a high-frequency TAE are simultaneously driven by REs [1,2]. All these modes are measured with the same toroidal mode number of n = 1 and propagate in the electron diamagnetic drift direction. An HFM is additionally identified, of which the frequency and temporal evolution are closely correlated with those of the TAE. Triplets of modes are identified whose frequencies and mode numbers, respectively, satisfy energy and momentum conservation laws, suggesting that three-wave coupling occurs [3]. This is confirmed through bicoherence analysis, which establishes unambiguously that the coupled modes have the requisite fixed phase relationships.
Further analysis reveals that the dominant HFM number from mode coupling depends on the BAE's position relative to the TAE: it is n = 0 when the BAE is outside, and n = 2 when it is inside. The nonlinear coupling equation can be derived from the vorticity equation and the quasi-neutrality condition. The Reynolds stresses, which are additive in this framework, contribute to the excitation of the HFM.

Reference:
1 Luo C. et al 2025 Nucl. Fusion 65 076020
[2] Zhu X. et al 2024 Nucl. Fusion 64 126023
[3] Zhu X. et al 2022 Phys. Plasmas 29 062504

Speaker: chenxi luo

112 Nonlinear dynamics of resistive hose modes for runaway electron beams in post-disruption tokamak plasmas

High-energy relativistic runaway electron (RE) beams emerge during plasma disruptions in tokamaks through avalanche multiplication processes. This phenomenon arises from the synergistic interplay of Dreicer acceleration and knock-on collisions, potentially jeopardizing plasma-facing components through localized energy deposition exceeding 10 MJ/m². Advanced MHD simulations have revealed that in post-thermal-quench RE scenarios with enhanced plasma resistivity (η > 10⁻³ Ω·m), the resistive hose modes, a kink-type instability characteristic of self-pinched relativistic electron beams, exhibit growth rates exceeding those of resistive tearing modes (TM) by up to two orders of magnitude at most. This disparity in instability development timescales poses significant challenges for beam confinement mitigation strategies in ITER-relevant disruption conditions. Moreover, the resistive hose mode may nonlinearly couple with other MHD modes, largely complicating the RE mitigating strategies. Thus, it is essential to investigate the nonlinear dynamics of resistive hose modes for runaway electron beams in post-disruption tokamak plasmas for its better mitigation. In this work, the macroscale effects of REs in tokamaks are treated as a separate cold beam-like fluid species in extended MHD simulations. The RE beam is considered as a source of resistance-free current density whose direction depends on the time evolving magnetic field that interacts with background plasma governed by a set of reduced MHD equations. Only the fraction of current carried by the bulk plasma is affected by resistivity. The linear and nonlinear properties of resistive hose mode are systematically investigated under different conditions. It is found that the resistive hose mode can saturate in a turbulent state. The nonlinear dynamics have all been discussed in detail. The nonlinear coupling between the resistive hose mode and TM is also discussed.

Speaker: Tong Liu (Dalian University of Technology)

10:50 Coffee break

Physics of EP modes and transport

113 A workflow for testing energetic particle transport models - from experiments to model validation

Projecting fast ion dynamics from present devices to a fusion reactor requires a combination of experimental results and numerical models. Confidence in the models relies on their extensive validation to ensure that the underlying physics is correct, as well as to assess their range of validity and possible limitations. Experiments on TCV have revisited the physics of well-known instabilities such as fishbones to gather a database for model validation. In parallel, several existing analysis and modeling tools have been combined in a single workflow to enable comprehensive tests and validation. Experimentally, the effects of fishbones on fast ions from Neutral Beam Injection (NBI) are characterized via diagnostics for confined and lost ions, namely neutron counters, Fast Ion D-Alpha (FIDA) and a Fast Ion Loss Detector (FILD) [1]. Mode properties are inferred from magnetic sensors and soft X-rays detectors covering the poloidal cross-section. Experimental data are used to guide modeling of the fast ion transport and loss induced by fishbones. Modeling is conducted via the particle-following code Orbit and with the combined TRANSP+kick model [2]. Results are in qualitative agreement with measurements from FIDA and FILD, indicating a redistribution of particles from the core – where the fishbone amplitude peaks – to mid-radius. Depending on the mode amplitude, fast ions near the co-passing/lost and trapped/lost boundaries can reach FILD, with characteristic values of energy and pitch that vary in time as the mode frequency and amplitude evolve [1]. The possibility to reconstruct modifications of the fast ion distribution induced by fishbones directly from experimental FIDA and FILD data is explored, by both forward modeling [1] and inversion techniques [3]. The results from modeling and forward/inverse analysis will be compared to assess the ability to infer the fast ion distribution response to the instability, as well as to assess the limitations of each method. The possible extension of those methods to more complex instabilities such as Alfvénic modes on TCV [4] will also be discussed.

[1] Q. Kernel et al., Experimental characterization of fishbones and associated energetic particle transport via correlated FIDA and FILD measurements on the TCV tokamak (this conference).
[2] M. Podestà et al., Plasma Phys. Control. Fusion 59 (2017) 095008.
[3] M. Rud et al., Resonant mode-particle interaction priors for energetic particle phase-space tomography (this conference).
[4] A. Jansen van Vuuren et al., Experimental investigations of Alfvén eigenmode control in TCV fast-ion plasmas (this conference).

Speaker: Mario Podesta

114 Experimental characterization of fishbones and associated energetic particle transport via correlated FIDA and FILD measurements on the TCV tokamak

Energetic-particle-driven instabilities can induce strong fast-ion redistribution and losses, thereby degrading plasma heating and confinement. This is a key concern for future burning plasmas such as ITER, where fusion-born alpha particles are expected to drive resonant wave-particle interactions. To limit these, it is important to understand the associated phase-space transport [1] and, more generally, the interplay between fast ions and MHD instabilities, to ultimately improve predictive capabilities [2] and developp robust control strategies for future reactors [3]. This work integrates multiple diagnostics to experimentally characterize fishbone-induced fast-ion transport in Neutral Beam Injection heated plasmas on the TCV tokamak across plasma shaping scans. Fishbone activity is identified from magnetic probes, electron cyclotron emission (ECE) and soft X-ray (SXR) diagnostics, providing the mode properties and timing of the instability cycle. To investigate the fast-ion response, event-synchronized conditional sampling and averaging over a large number of similar fishbone bursts is applied to Fast-Ion Deuterium-Alpha (FIDA) spectroscopy and Fast-Ion Loss Detector (FILD) measurements [4, 5] and combined with inversion techniques [6–8]. Confined and lost fast-ion velocity-space distributions are reconstructed and correlated throughout a typical fishbone cycle. Following the instability onset, reduced levels of FIDA emission are observed across the full wavelength range, suggesting a broad fast-ion redistribution over the measured energy domain. The depletion is stronger at higher energies and in the plasma core where the mode is localized. The dual-camera FILD system resolves the temporal velocity-space evolution of fast-ion losses. The signals exhibit coherent frequency signatures consistent with the activity observed in magnetic and ECE diagnostics. The velocity-space reconstructions show two distinct pitch-angle loss structures, with intense high-energy losses occurring near the mode onset, followed by a progressive shift toward lower energies and a decay of the loss amplitude. Together, these measurements reveal consistent signatures of core fast-ion redistribution and edge fast-ion losses, highlighting an energy-dependent transport process during fishbone cycles associated with resonant wave-particle interactions. This study establishes a general experimental framework for a systematic reconstruction of fast-ion redistribution and losses during MHD activity and supports the validation of predictive fast-ion transport models for future fusion devices, in different plasma configurations, including negative triangularity scenarios [9].

[1] Heidbrink W. W. Maxwell Prize lecture on measuring the phase-space transport of energetic particles in magnetic fusion experiments Physics of Plasmas 33 (2026) 030601
[2] Podestà M. et al A workflow for testing energetic particle transport models - from experiments to model validation (this conference)
[3] Jansen van Vuuren A. et al Experimental investigations of Alfvén eigenmode control in TCV fast-ion plasmas (this conference)
[4] Geiger B. et al Fast-ion transport in low density L-mode plasmas at TCV using FIDA spectroscopy and the TRANSP code Plasma Physics and Controlled Fusion 59 (2017) 115002
[5] Poley-Sanjuán J. et al First microsecond velocity-space resolved simultaneous measurements of co-andcounter-current fast-ion losses in forward and reverse magnetic field in a tokamak Nuclear Fusion 65 (2025) 076006
[6] Rud M. et al Bayesian velocity-space tomography with collision- and charge-exchange-physics prior from fast-ion D-alpha measurements at TCV with uncertainty quantification Plasma Physics and Controlled Fusion 68 (2026) 015035
[7] Rud M. et al Resonant mode-particle interaction priors for energetic particle phase-space tomography (this conferencee)
[8] Galdon-Quiroga J. et al Velocity-space sensitivity and tomography of scintillator-based fast-ion loss detectors Plasma Physics and Controlled Fusion 60 (2018) 105005
[9] Poley-Sanjuán J. et al Fast ion confinement in negative triangularity plasmas on the TCV tokamak Nuclear Fusion 66 (2026) 046012

Speaker: Mr Quentin Kernel (EPFL - Swiss Plasma Center)

115 Vertical Displacement Oscillatory Modes in Tokamak Plasmas

An overview on Vertical Displacement Oscillatory Modes driven by fast ions is presented. These modes have been identified by analytic work in 2022 [1,2] and their existence for realistic tokamak geometry has been confirmed by numerical simulations using the NIMROD code [3]. Together with Global Alfvén Eigenmodes (GAE), VDOM are a candidate to explain the observation of axisymmetric, saturated magnetic fluctuations, with toroidal mode number n=0, in recent JET [4] and TCV [5] experiments. In this presentation, we shall focus on the important differences between GAE and VDOM as clarified by semi-analytic work in the cylindrical (straight tokamak) limit. VDOM are global, external plasma modes whose existence rely on the interaction between the plasma current and the currents induced on the nearby conducting wall when the plasma boundary is perturbed. An analytic dispersion relation that contains both VDOM and ideal-MHD unstable vertical displacements is discussed. While VDOM are generally immune against continuum damping, if the wall is brought very close to the plasma boundary and the plasma density remains relatively large up to the last-closed-flux-surface (as in the case of H-mode discharges), the VDOM oscillation frequency may interact with the Alfvén continuum spectrum, giving rise to continuum damping. Specific types of fast ion distribution functions that can provide an instability drive for n=0 modes are considered. With the fast ion distribution function expressed as f_h (ℇ,Λ,P_φ) (ℇ is the energy, Λ=μB_0/ℇ is a pitch angle variable, where μ is the magnetic momentum, and P_φ is the toroidal angular momentum), the drive of axisymmetric modes requires either a non-monotonic distribution in the energy direction (with Λ and P_φ held constant) or anisotropy, i.e. a variation of f_h with Λ. For fusion alpha particles, it is shown that finite orbit width effects lead to an anisotropy with a peak of (∂f_h)⁄∂Λ in the region of trapped orbits. A possible drive due this anisotropy is discussed. Numerical simulations of n=0 modes obtained with the NIMROD code are shown, including a preliminary assessment of the fast ion instability drive for these modes.

[1] T. Barberis, F. Porcelli, and A. Yolbarsop, Fast ion driven vertical modes in magnetically confined toroidal plasmas, Nuclear Fusion Letter 62, 064002 (2022).
[2] T. Barberis, A. Yolbarsop, and F. Porcelli, Vertical displacement oscillatory modes in tokamak plasma, Journal of Plasma Physics 88, 905880511 (2022).
[3] T. Barberis, C.C. Kim, F. Porcelli, D. Banerjee, N. Hawkes, Ye O. Kazakov, Y.Q. Liu, H.J.C. Oliver, S.E. Sharapov, A. Yolbarsop, NIMROD Team and JET Contributors, Simulations of vertical displacement oscillatory modes and global Alfvén Eigenmodes in JET geometry, Nuclear Fusion 64, 126064 (2024).
[4] H.J.C. Oliver, D. King, Ž. Štancar, S.E. Sharapov, D. Banerjee, T. Barberis, R. Coelho, I. Coffey, M. Dreval, M. Fitzgerald, L. Frassinetti, C. Giroud, N. Hawkes, D. Keeling, C.C. Kim, E. Lerche, F. Porcelli, G. Szepesi, Axisymmetric eigenmodes excited by alpha particle energy gradients in JET D-T plasmas, Physical Review Letters, 136 (2026) 055101.
[5] M. Dreval, S.E. Sharapov, H.J.C. Oliver, M. Fitzgerald, A.N. Karpushov, A. Jansen van Vuuren, J. Poley, M. Podesta, F. Porcelli, the TCV Team and the EUROfusion Tokamak Exploitation Team, Axisymmetric global Alfvén eigenmodes in the TCV tokamak, under review for publication in Nuclear Fusion (2026).

Speaker: Francesco Porcelli (Politecnico di Torino)

12:45 Lunch break

Panel discussion

Convener: William Heidbrink

Closing remarks and Announcements

Convener: Mario Podesta (SPC/EPFL, Lausanne)