In a tokamak, a toroidal device for magnetic confinement fusion, the plasma edge represents the transition region between the thermonuclear fusion core and the solid wall of the vacuum chamber. The edge plasma is a complex and articulated physical system, where multi-scale transport phenomena, atomic processes, and plasma–surface interactions are tightly coupled. These mechanisms jointly determine the efficiency of power exhaust and the resulting thermal and particle loads on plasma-facing components, whose characterisation and control are critical for the feasibility and safe operation of future fusion reactors. In this context, numerical modelling plays a fundamental role, complementing experimental investigations in both a descriptive and predictive sense. On the one hand, modelling provides an interpretative framework for existing experimental data; on the other, it enables the prediction of plasma behaviour in operational regimes that are difficult to access experimentally, as well as the design of future scenarios. Among the major codes for edge plasma studies, SOLPS-ITER, adopted in this thesis, has emerged as the state-of-the-art modelling tool of choice for ITER simulations and DEMO design. The objective of this thesis is twofold. First, to numerically investigate fusion-relevant edge plasma scenarios that remain little explored — either due to their unconventional plasma composition or their non-standard magnetic geometry — in order to gain physical insight. Second, to broaden the range of applicability of state-of-the-art modelling tools beyond their most common use cases, assessing their robustness and limitations across a variety of devices and configurations. The first part of the work focuses on linear plasma devices, through a systematic modelling campaign of helium and argon plasmas in the GyM machine. For helium plasmas, the novel explicit inclusion of metastable states within SOLPS-ITER was demonstrated, showing their impact on plasma density and temperature profiles. The generation of a validated background plasma opened the way to coupling with erosion modelling codes such as ERO2.0, enabling an integrated description of plasma behaviour and surface erosion in dedicated plasma-wall interaction experiments. In the case of argon plasmas, motivated by cross-machine experiments, SOLPS-ITER simulations were optimised against Langmuir probe data and compared with a simplified 0D global model, highlighting the predictive capabilities of reduced approaches. The second part addresses tokamak configurations of reactor relevance. In TCV, the role of magnetic geometry was investigated by comparing L-mode discharges with positive and negative upper triangularity, while keeping divertor geometry and heating power fixed. The analysis reproduced an attached and hotter divertor in negative triangularity configurations—under the assumption of suppressed cross-field transport—and shed light on neutral dynamics at the divertor, identifying access to detachment as a critical aspect to assess the viability of negative triangularity as a reactor solution. In ASDEX Upgrade, a pure helium H-mode plasma was modelled with SOLPS-ITER and validated against experimental data. The study distinguished the contributions of different helium charge states to the divertor particle flux, providing the basis for eventual erosion modelling studies. This work extends the applicability of edge plasma modelling to non-hydrogenic H-mode plasmas, which remain largely unexplored but are directly relevant to ITER, where helium will inevitably accumulate as fusion ash. Overall, this thesis provides new validated insights into boundary plasma behaviour in both linear and toroidal devices, while broadening the scope of numerical tools for edge plasma and PMI studies. The results demonstrate that state-of-the-art modelling frameworks can successfully reproduce and interpret unconventional edge plasma scenarios, and can serve as predictive tools for the design of future experiments and fusion devices. In this way, the work contributes to consolidating the role of numerical modelling as a crucial pillar, alongside experiment and theory, in the pursuit of nuclear fusion as a sustainable energy source.
In un tokamak, macchina toroidale per la fusione a confinamento magnetico, il plasma di bordo rappresenta la regione di transizione tra il core termonucleare e la parete solida della camera da vuoto. Il plasma di bordo costituisce un sistema fisico complesso e articolato, in cui fenomeni di trasporto su più scale, processi atomici e interazioni plasma–superficie risultano strettamente accoppiati. Questi meccanismi determinano congiuntamente l’efficienza dello smaltimento di potenza e i conseguenti carichi termici e di particelle sulla prima parete, la cui caratterizzazione e gestione sono aspetti cruciali per la fattibilità e il funzionamento sicuro dei futuri reattori a fusione. In questo contesto, la modellistica numerica riveste un ruolo fondamentale, integrando le indagini sperimentali sia in chiave descrittiva sia in chiave predittiva. Da un lato, infatti, fornisce un quadro interpretativo dei dati esistenti; dall’altro permette di predire il comportamento del plasma in condizioni operative difficilmente accessibili sperimentalmente e di contribuire alla progettazione di nuovi scenari di plasma. Tra i principali codici per lo studio del plasma di bordo, SOLPS-ITER, adottato in questa tesi, si è affermato come lo strumento di riferimento per il design di ITER e DEMO. L’obiettivo di questo lavoro è duplice. Da un lato, investigare numericamente scenari di plasma di bordo rilevanti per la fusione ma ancora poco esplorati, sia per la loro composizione non convenzionale, sia per la geometria magnetica atipica, al fine di acquisire una più profonda comprensione fisica. Dall’altro, estendere il campo di applicazione degli strumenti modellistici di ultima generazione oltre i casi d’uso più consolidati, valutandone robustezza e limiti. La prima parte del lavoro riguarda le macchine lineari, con una campagna di simulazioni sistematiche di plasmi di elio e di argon nella macchina GyM. Per i plasmi di elio, è stata dimostrata l’inclusione esplicita degli stati metastabili all’interno di SOLPS-ITER, mostrando il loro impatto sui profili di densità e temperatura. La generazione di un plasma di background validato ha aperto la strada all’accoppiamento con codici di simulazione dell’erosione, come ERO2.0, consentendo una descrizione integrata del comportamento del plasma e dei fenomeni di erosione superficiale in esperimenti dedicati di interazione plasma–parete. Nel caso dei plasmi di argon, studiati nell’ambito di esperimenti multi-macchina, le simulazioni SOLPS-ITER sono state ottimizzate rispetto ai dati delle sonde di Langmuir e confrontate con un modello globale 0D semplificato, mettendo in evidenza le capacità predittive degli approcci ridotti. La seconda parte affronta configurazioni di tokamak rilevanti. In TCV è stato analizzato il ruolo della geometria magnetica confrontando scariche in L-mode con triangolarità positiva e negativa, ma identiche geometria del divertore e potenza di riscaldamento. L’analisi ha riprodotto un divertore in regime attaccato e più caldo per la configurazione a triangolarità negativa—assumendo un trasporto radiale ridotto—e ha chiarito la dinamica dei neutri al divertore, identificando l’accesso al detachment come aspetto cruciale per valutare la triangolarità negativa come soluzione per un futuro reattore. In ASDEX Upgrade è stato modellato con SOLPS-ITER, e validato contro dati sperimentali, un plasma di puro elio in H-mode. Lo studio ha permesso di distinguere i contributi dei diversi stati di carica dell’elio al flusso di particelle al divertore, gettando le basi per successivi studi di erosione. Questo lavoro estende l’applicabilità della modellistica del plasma di bordo a plasmi in H-mode non idrogenici, scenari ancora poco esplorati, ma direttamente rilevanti per ITER, dove l’elio si accumulerà inevitabilmente come prodotto di fusione. Complessivamente, questa tesi fornisce nuovi risultati validati sul comportamento del plasma di bordo in dispositivi lineari e toroidali, ampliando al contempo l’ambito di applicazione degli strumenti numerici per lo studio del plasma di bordo e dell'interazione plasma-parete. I risultati dimostrano che i codici di riferimento possono riprodurre e interpretare con successo scenari non convenzionali e che possono costituire strumenti predittivi per la progettazione di futuri esperimenti e reattori a fusione. In questo modo, il lavoro contribuisce a consolidare il ruolo della modellistica numerica come pilastro essenziale, accanto all’esperimento e alla teoria, nello sviluppo della fusione nucleare come fonte di energia sostenibile.
Numerical modelling of the boundary plasma in magnetic fusion devices
Mombelli, Fabio
2025/2026
Abstract
In a tokamak, a toroidal device for magnetic confinement fusion, the plasma edge represents the transition region between the thermonuclear fusion core and the solid wall of the vacuum chamber. The edge plasma is a complex and articulated physical system, where multi-scale transport phenomena, atomic processes, and plasma–surface interactions are tightly coupled. These mechanisms jointly determine the efficiency of power exhaust and the resulting thermal and particle loads on plasma-facing components, whose characterisation and control are critical for the feasibility and safe operation of future fusion reactors. In this context, numerical modelling plays a fundamental role, complementing experimental investigations in both a descriptive and predictive sense. On the one hand, modelling provides an interpretative framework for existing experimental data; on the other, it enables the prediction of plasma behaviour in operational regimes that are difficult to access experimentally, as well as the design of future scenarios. Among the major codes for edge plasma studies, SOLPS-ITER, adopted in this thesis, has emerged as the state-of-the-art modelling tool of choice for ITER simulations and DEMO design. The objective of this thesis is twofold. First, to numerically investigate fusion-relevant edge plasma scenarios that remain little explored — either due to their unconventional plasma composition or their non-standard magnetic geometry — in order to gain physical insight. Second, to broaden the range of applicability of state-of-the-art modelling tools beyond their most common use cases, assessing their robustness and limitations across a variety of devices and configurations. The first part of the work focuses on linear plasma devices, through a systematic modelling campaign of helium and argon plasmas in the GyM machine. For helium plasmas, the novel explicit inclusion of metastable states within SOLPS-ITER was demonstrated, showing their impact on plasma density and temperature profiles. The generation of a validated background plasma opened the way to coupling with erosion modelling codes such as ERO2.0, enabling an integrated description of plasma behaviour and surface erosion in dedicated plasma-wall interaction experiments. In the case of argon plasmas, motivated by cross-machine experiments, SOLPS-ITER simulations were optimised against Langmuir probe data and compared with a simplified 0D global model, highlighting the predictive capabilities of reduced approaches. The second part addresses tokamak configurations of reactor relevance. In TCV, the role of magnetic geometry was investigated by comparing L-mode discharges with positive and negative upper triangularity, while keeping divertor geometry and heating power fixed. The analysis reproduced an attached and hotter divertor in negative triangularity configurations—under the assumption of suppressed cross-field transport—and shed light on neutral dynamics at the divertor, identifying access to detachment as a critical aspect to assess the viability of negative triangularity as a reactor solution. In ASDEX Upgrade, a pure helium H-mode plasma was modelled with SOLPS-ITER and validated against experimental data. The study distinguished the contributions of different helium charge states to the divertor particle flux, providing the basis for eventual erosion modelling studies. This work extends the applicability of edge plasma modelling to non-hydrogenic H-mode plasmas, which remain largely unexplored but are directly relevant to ITER, where helium will inevitably accumulate as fusion ash. Overall, this thesis provides new validated insights into boundary plasma behaviour in both linear and toroidal devices, while broadening the scope of numerical tools for edge plasma and PMI studies. The results demonstrate that state-of-the-art modelling frameworks can successfully reproduce and interpret unconventional edge plasma scenarios, and can serve as predictive tools for the design of future experiments and fusion devices. In this way, the work contributes to consolidating the role of numerical modelling as a crucial pillar, alongside experiment and theory, in the pursuit of nuclear fusion as a sustainable energy source.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD_thesis_mombelli-politesi.pdf
accessibile in internet per tutti
Descrizione: Manuscript
Dimensione
18.94 MB
Formato
Adobe PDF
|
18.94 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/249917