Plasma-material interaction (PMI) is among the main challenges for the future exploitation of nuclear fusion as a reliable energy source. Both plasma and materials are affected by the interaction between these two opposite states of matter. The high particle and power fluxes expected in future devices could cause severe damage to materials, such as erosion due to e.g. physical sputtering, and morphology changes, which can influence their thermo-mechanical properties. Erosion, in turn, represents an impurity source for the plasma, thus increasing dilution and radiation cooling, potentially preventing high energy production. Finally, heavy impurities can re/co-deposit with fuel species in different locations inside the device, possibly changing surface properties of underneath materials and increasing radioactive tritium inventory. The unavoidable presence of helium (He) in plasma as a product of D-T reaction could influence and enhance these phenomena due to its peculiar interaction with materials. Being able to predict and possibly mitigate these effects is thus of paramount importance. PMI phenomena are often studied in tokamaks. However, devices fully comparable to reactor relevant conditions are still not available and their complex environment often hinders the complete understanding of experimental dynamics. For these reasons, parallel activities in simpler experiments and the availability of modelling tools could significantly support tokamak research, with the possibility of extrapolating present findings to future devices, such as the DTT tokamak. On the one hand, linear plasma devices (LPDs) offer a complementary and cost-effective alternative to study PMI expected in future tokamaks, exposing suitable samples to well-defined and repeatable plasma conditions. On the other one, the 3D Monte-Carlo code ERO2.0 is widely recognized as state-of-the-art for erosion and impurity migration studies. The main goal of this PhD thesis thus consists in supporting the interpretation of experiments carried out in LPDs and tokamaks by means of ERO2.0 modelling, eventually exploiting the acquired experience to make reliable predictions for DTT design. For this purpose, erosion and impurity migration phenomena are studied in different devices, from the microscale to machine-scale global simulations. Tungsten (W) coatings with different morphology have been deposited on substrates with different roughness and exposed to He plasma in the linear device GyM, to study topography effect on sample erosion. The microscale topography of exposed samples has also been directly imported in ERO2.0 simulations for comparison. Both experiments and simulations agree in identifying the mean surface inclination angle as a better descriptor of sputtering yield variation with respect to roughness. SOLPS-ITER and ERO2.0 codes have been coupled for the first time in linear geometry to perform global modelling of wall erosion and transport in GyM. In particular, wall eroded impurities are shown to contribute to sample erosion in range 15 − 20% for impact He energy in proximity of the sputtering threshold. ERO2.0 modelling has also been applied in tokamak geometry to support experiment interpretation in AUG D and He plasma discharges. In the former case, global modelling of first wall erosion and trasport to the divertor region shows that wall eroded W could contribute to gold markers erosion in range 5−15% according to free simulation parameters, while it drops below 1% for Mo ones. In He discharges, the role of He ion concentration on divertor erosion has been studied by means of a simple analytical model and with ERO2.0, showing the dominance of He2+ erosion in this plasma temperature regime. A good agreement is found with experimental data, considering a variation of plasma parameters within their uncertainty and including a reasonable presence of light impurities in plasma. Finally, SOLPS-ITER and ERO2.0 W density predictions in plasma have been compared, highlighting advantages and limitations of the two models. This experience has been exploited to provide first indications for DTT design. Two lower single-null scenarios have been investigated, i.e. attached plasma in pure D and detached with Ne seeding, with the aim of assessing the erosion/deposition pattern and Wcore contamination from the divertor region. A comparable peak erosion is observed in the two scenarios in ELM-free conditions, while net erosion is found to be overall dominated by ELMs. Wcore contamination is observed to stay below acceptable limits in all the investigated scenarios. The horizontal flat parts of both targets and the central flat part of the dome are identified as the least screened areas of DTT divertor, which could raise concerns in case of aWsource there. Thanks to the comparison with WEST experience, the DTT good screening properties are ascribed to a combination of closed divertor geometry and to the high electron density near surface.
L’interazione plasma-materiali (PMI) è una delle principali sfide da affrontare per essere in grado di sfruttare, in futuro, la fusione nucleare come sorgente affidabile di energia. Sia il plasma che i materiali sono influenzati dall’interazione fra questi due stati opposti della materia. Gli alti flussi di particelle e potenza attesi nei reattori futuri possono causare danni importanti ai materiali, come l’erosione in seguito a processi di sputtering fisico, e cambiamenti morfologici, che possono influenzare le proprietà termo-meccaniche dei materiali stessi. Dall’altro lato, l’erosione rappresenta una sorgente di impurezze per il plasma, causando diluizione e raffreddamento radiativo, arrivando potenzialmente a impedire la produzione di grandi quantità di energia. Infine, le impurezze pesanti posso ridepositarsi con le specie del combustibile in diverse aree all’interno della macchina, andando a modificare le proprietà del materiale sottostante e aumentando l’inventario di trizio radiattivo ritenuto nel materiale. L’inevitabile presenza di elio (He) nel plasma a seguito delle reazioni D-T può ulteriormente cambiare e rafforzare questi fenomeni a causa della sua particolare interazione coi materiali. Essere quindi in grado di predire e possibilmente mitigare questi effetti è fondamentale. I fenomeni di PMI sono spesso studiati nei tokamak. Tuttavia, dispositivi aventi caratteristiche confrontabili con quelle attese nei reattori futuri non sono al momento disponibili e il loro ambiente complesso spesso impedisce una completa comprensione della dinamica sperimentale. Per queste ragioni, attività parallele in esperimenti più semplici e la disponibilità di strumenti modellistici potrebbero supportare in modo significativo la ricerca nei tokamak, con la possibilità anche di estrapolare i risultati odierni verso macchine future come DTT. Da un lato, le macchine lineari offrono un’alternativa complementare ed economica per studiare la PMI attesa in tokamak futuri, esponendo campioni opportuni a condizioni di plasma ben definite e riproducibili. Dall’altro lato, il codice 3D Monte-Carlo ERO2.0 è ampiamente riconosciuto come lo stato dell’arte per studi di erosione e migrazione di impurezze. Il principale obiettivo di questa tesi di dottorato consiste quindi nel supportare l’interpretazione di esperimenti compiuti in macchine lineari e tokamak attraverso il codice ERO2.0, sfruttando infine questa esperienza per fare predizioni affidabili per il design di DTT. A questo scopo, i fenomeni di erosione e migrazione di impurezze sono studiati in diverse macchine, dalla scala microscopica fino a simulazioni globali alla scala macchina. Film di tungsteno (W) con diverse morfologie sono stati depositati su substrati con diversa rugosità ed esposti a plasmi di elio nella macchina lineare GyM, per studiare l’effetto della topografia sull’erosione dei campioni. La topografia alla microscala dei campioni esposti è stata anche importata direttamente in ERO2.0 per confronto. Sia gli esperimenti che le simulazioni concordano nell’identificazione dell’angolo medio di inclinazione della superficie come miglior descrittore della variazione dell’erosione rispetto alla rugosità. I codici SOLPS-ITER ed ERO2.0 sono stati anche accoppiati per la prima volta in macchina lineare allo scopo di simulare l’erosione della parete e il trasporto in GyM. In particolare, si è dimostrato come le impurezze erose dalla parete possano contribuire all’erosione dei campioni nell’ordine del 15−20% in caso di energia degli ioni di He incidenti vicina alla soglia di sputtering. ERO2.0 è stato anche applicato in tokamak per supportare l’interpretazione di scariche di D e He in AUG. Nel primo caso, simulazioni globali di erosione della prima parete e trasporto verso il divertore mostrano come il W eroso dalla parete possa contribuire all’erosione dei campioni di oro posizionati nella regione del divertore nella misura del 5−15% a seconda dei parametri liberi di simulazione, mentre il contributo crolla sotto l’1% nel caso di campioni di Mo. Nelle scariche in He, il ruolo della concentrazione degli ioni di He sull’erosione del divertore è stato studiato attraverso un semplice modello analitico e con ERO2.0, dimostrando il ruolo predominante dell’He2+ in questo regime di temperature del plasma. Un buon accordo con i dati esperimentali è stato trovato considerando una variazione dei parametri di plasma all’interno della loro incertezza e includendo una ragionevole presenza di impurezze leggere nel plasma. Infine, le predizioni sulla densità delWnel plasma da SOLPS-ITER e ERO2.0 sono state confrontate, evidenziando vantaggi e limitazioni dei due modelli. Questa esperienza è stata sfruttata per fornire prime indicazioni per il design di DTT. Due scenari a singolo nullo inferiore sono stati investigati, vale a dire il caso di plasma attaccato in puro D e quello staccato con immissione di neon (Ne), con lo scopo di stimare la distribuzione dell’erosione e deposizione e la contaminazione del core dalla regione del divertore. L’erosione di picco nei due scenari in assenza di ELM è confrontabile, mentre l’erosione netta è dominata nel complesso dagli ELM stessi. La contaminazione del core rimane sotto valori accettabili in tutti gli scenari investigati. Le superfici piatte orizzontali dei due target e la superficie piatta centrale del dome sono identificate come le aree meno schermate del divertore di DTT, le quali potrebbero destare preoccupazioni in caso di sorgenti diWin quelle zone. Grazie al confronto con WEST, le buone proprietà di schermaggio di DTT sono spiegate attraverso la combinazione della geometria chiusa del divertore e dell’alta densità elettronica in prossimità della superficie.
Erosion and impurity migration modelling in linear devices and tokamaks
ALBERTI, GABRIELE
2023/2024
Abstract
Plasma-material interaction (PMI) is among the main challenges for the future exploitation of nuclear fusion as a reliable energy source. Both plasma and materials are affected by the interaction between these two opposite states of matter. The high particle and power fluxes expected in future devices could cause severe damage to materials, such as erosion due to e.g. physical sputtering, and morphology changes, which can influence their thermo-mechanical properties. Erosion, in turn, represents an impurity source for the plasma, thus increasing dilution and radiation cooling, potentially preventing high energy production. Finally, heavy impurities can re/co-deposit with fuel species in different locations inside the device, possibly changing surface properties of underneath materials and increasing radioactive tritium inventory. The unavoidable presence of helium (He) in plasma as a product of D-T reaction could influence and enhance these phenomena due to its peculiar interaction with materials. Being able to predict and possibly mitigate these effects is thus of paramount importance. PMI phenomena are often studied in tokamaks. However, devices fully comparable to reactor relevant conditions are still not available and their complex environment often hinders the complete understanding of experimental dynamics. For these reasons, parallel activities in simpler experiments and the availability of modelling tools could significantly support tokamak research, with the possibility of extrapolating present findings to future devices, such as the DTT tokamak. On the one hand, linear plasma devices (LPDs) offer a complementary and cost-effective alternative to study PMI expected in future tokamaks, exposing suitable samples to well-defined and repeatable plasma conditions. On the other one, the 3D Monte-Carlo code ERO2.0 is widely recognized as state-of-the-art for erosion and impurity migration studies. The main goal of this PhD thesis thus consists in supporting the interpretation of experiments carried out in LPDs and tokamaks by means of ERO2.0 modelling, eventually exploiting the acquired experience to make reliable predictions for DTT design. For this purpose, erosion and impurity migration phenomena are studied in different devices, from the microscale to machine-scale global simulations. Tungsten (W) coatings with different morphology have been deposited on substrates with different roughness and exposed to He plasma in the linear device GyM, to study topography effect on sample erosion. The microscale topography of exposed samples has also been directly imported in ERO2.0 simulations for comparison. Both experiments and simulations agree in identifying the mean surface inclination angle as a better descriptor of sputtering yield variation with respect to roughness. SOLPS-ITER and ERO2.0 codes have been coupled for the first time in linear geometry to perform global modelling of wall erosion and transport in GyM. In particular, wall eroded impurities are shown to contribute to sample erosion in range 15 − 20% for impact He energy in proximity of the sputtering threshold. ERO2.0 modelling has also been applied in tokamak geometry to support experiment interpretation in AUG D and He plasma discharges. In the former case, global modelling of first wall erosion and trasport to the divertor region shows that wall eroded W could contribute to gold markers erosion in range 5−15% according to free simulation parameters, while it drops below 1% for Mo ones. In He discharges, the role of He ion concentration on divertor erosion has been studied by means of a simple analytical model and with ERO2.0, showing the dominance of He2+ erosion in this plasma temperature regime. A good agreement is found with experimental data, considering a variation of plasma parameters within their uncertainty and including a reasonable presence of light impurities in plasma. Finally, SOLPS-ITER and ERO2.0 W density predictions in plasma have been compared, highlighting advantages and limitations of the two models. This experience has been exploited to provide first indications for DTT design. Two lower single-null scenarios have been investigated, i.e. attached plasma in pure D and detached with Ne seeding, with the aim of assessing the erosion/deposition pattern and Wcore contamination from the divertor region. A comparable peak erosion is observed in the two scenarios in ELM-free conditions, while net erosion is found to be overall dominated by ELMs. Wcore contamination is observed to stay below acceptable limits in all the investigated scenarios. The horizontal flat parts of both targets and the central flat part of the dome are identified as the least screened areas of DTT divertor, which could raise concerns in case of aWsource there. Thanks to the comparison with WEST experience, the DTT good screening properties are ascribed to a combination of closed divertor geometry and to the high electron density near surface.File | Dimensione | Formato | |
---|---|---|---|
PhD thesis - Alberti_revised.pdf
accessibile in internet per tutti a partire dal 16/10/2025
Descrizione: PhD thesis
Dimensione
115.96 MB
Formato
Adobe PDF
|
115.96 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/10589/229092