This thesis work presents the results of computational fluid dynamics (CFD) simulations of bubbly flows in a bubble column reactor, with particular focus on resolution-adaptive modeling and morphological transitions from dispersed unresolved bubbles to continuous fully resolved ones. Transient 3D simulations were based on an Euler–Euler approach and employing the MultiMorph model, a multifield morphology adaptive multifluid approach developed at the Helmholtz-Zentrum Dresden-Rossendorf (HZDR). The methodology enables a dynamic transition between dispersed and resolved interface representations depending on the local mesh resolution. This hybrid strategy allows the simultaneous treatment of small, dispersed bubbles and larger resolved gas structures within a single, physically consistent framework. The investigation is based on a set of established closure relations for interphase momentum exchange among dispersed and continuous phases, including drag, lift, wall lubrication, virtual mass and turbulent dispersion forces. The dispersed phase was modeled using a poly-dispersed Population Balance Modeling (PBM) approach based on the Method of Classes for polydispersed regimes. Both two and three class representations were evaluated to model the coexistence of small, medium and large bubbles and their impact on the radial void fraction profiles. The results showed that this approach is necessary to accurately capture the physical phenomena occurring in the system, particularly the tran sition between homogeneous and heterogeneous regimes, which is governed by the sign of the lift coefficient and the bubble size. Particular attention is devoted to the implementation and extension of a clustering method ology for modeling the morphology-transfer, the Over-Resolved To Continuous (OTC) method, enabling the identification and conversion of over-resolved dispersed gas regions into continuous resolved bubbles, ensuring mass and momentum conservation during the transitions. The simulations demonstrate the capability of the MultiMorph framework to consistently represent both dispersed and resolved gas structures within the same com putational domain, maintaining numerical stability across mesh refinement levels. The proposed framework is first verified through a mesh convergence study and dedicated test cases, assessing numerical robustness and consistency across morphology transitions. Subsequently, the model is validated against experimental data from the literature, considering both monodispersed and polydispersed configurations and different mesh resolutions. Global quantities such as gas holdup and local features, including radial profiles of gas volume fraction, are analyzed and compared with experi mental measurements. The results show that the polydispersed formulation significantly improves the prediction of radial profiles. Results demonstrate that the morphology-adaptive approach successfully captures regime transitions, bubble clustering phenomena and the interactions between dispersed and resolved structures, while maintaining numerical stability. The integration of PBM further improves the predictive capability in polydispersed regimes. The study confirms the potential of the MultiMorph framework as a robust and scalable tool for advanced bubble column simulations and represents a foundational step toward the modeling of more complex industrial multiphase systems.
Questo lavoro di tesi illustra i risultati di simulazioni CFD di flussi a bolle in un reattore a colonna a bolle verticale, con un’attenzione particolare allo sviluppo di un modello adattivo basato sulla risoluzione della mesh per lo studio delle transizioni morfologiche da strutture gassose disperse non risolte a bolle continue con interfacce completamente risolte. Le simulazioni 3D presentano un approccio Euleriano-Euleriano multifase basato sul modello MultiMorph, un metodo adattivo a due fluidi sviluppato all’Helmholtz-Zentrum Dresden Rossendorf (HZDR), che permette di definire transizioni dinamiche tra rappresentazioni di interfacce disperse e risolte, calibrate sulla risoluzione locale della griglia. Questa strategia ibrida consente di modellare simultaneamente piccole bolle disperse e grandi risolte all’interno di un unico framework fisicamente coerente. L’analisi impiega un set consolidato di modelli di chiusura per gli scambi di momento tra fasi disperse e continue, includendo le forze di drag, lift, wall force, virtual mass e turbulent dispersion. La fase dispersa è stata simulata attraverso l’introduzione di equazioni addizionali per descrivere il PBM, basato sul Metodo delle Classi nei regimi polidispersi. Sono state testate rappresentazioni a due e tre classi per simulare la coesistenza di bolle piccole, medie e grandi, valutandone l’impatto sui profili radiali di frazione di vuoto. I risultati confermano la necessità di questo approccio per rappresentare fedelmente i fenomeni fisici del sistema, soprattutto per quanto riguarda la transizione tra regimi omogeneo ed eterogeneo, determinata dal segno del coefficiente di lift e, dunque, dalle dimensioni delle bolle stesse. Un’attenzione particolare è rivolta all’implementazione e all’ampliamento di un algoritmo di clustering per il trasferimento di fase tra le diverse morfologie presenti nel sistema, il metodo definito Over-Resolved To Continuous (OTC), che identifica e converte regioni gassose disperse eccessivamente risolte in bolle continue, preservando massa e momento durante tali transizioni. Le simulazioni evidenziano come il framework MultiMorph sia capace di rappresentare in modo stabile e coerente sia le strutture disperse che quelle risolte nello stesso dominio. Il modello è stato prima verificato tramite uno studo sulla convergenza della mesh e successivamente utilizzando test mirati per valutare robustezza numerica e coerenza fisica nelle transizioni morfologiche. È seguito il confronto con dati sperimentali presenti in letteratura, esaminando configurazioni monodisperse e polidisperse a varie risoluzioni. Sono state analizzate grandezze globali, come l’holdup di gas, e locali quali i profili radiali. La modellazione polidispersa ha infine dimostrato un netto miglioramento nella previsione dei profili radiali. In sintesi, l’approccio introdotto in questa tesi permette di modellare efficacemente transizioni di regime, clustering di bolle e interazioni tra strutture disperse e risolte, garantendo stabilità numerica. L’integrazione delle PBM, inoltre, potenzia le prestazioni nei regimi polidispersi. Questo studio valida il framework MultiMorph come strumento robusto e potenzialmente scalabile per simulazioni più complesse di colonne a bolle, aprendo la strada alla modellazione di sistemi multifase industriali.
Morphology-adaptive simulations of bubbly flow in poly-dispersed regimes
Tagliabue, Elena Antonia
2024/2025
Abstract
This thesis work presents the results of computational fluid dynamics (CFD) simulations of bubbly flows in a bubble column reactor, with particular focus on resolution-adaptive modeling and morphological transitions from dispersed unresolved bubbles to continuous fully resolved ones. Transient 3D simulations were based on an Euler–Euler approach and employing the MultiMorph model, a multifield morphology adaptive multifluid approach developed at the Helmholtz-Zentrum Dresden-Rossendorf (HZDR). The methodology enables a dynamic transition between dispersed and resolved interface representations depending on the local mesh resolution. This hybrid strategy allows the simultaneous treatment of small, dispersed bubbles and larger resolved gas structures within a single, physically consistent framework. The investigation is based on a set of established closure relations for interphase momentum exchange among dispersed and continuous phases, including drag, lift, wall lubrication, virtual mass and turbulent dispersion forces. The dispersed phase was modeled using a poly-dispersed Population Balance Modeling (PBM) approach based on the Method of Classes for polydispersed regimes. Both two and three class representations were evaluated to model the coexistence of small, medium and large bubbles and their impact on the radial void fraction profiles. The results showed that this approach is necessary to accurately capture the physical phenomena occurring in the system, particularly the tran sition between homogeneous and heterogeneous regimes, which is governed by the sign of the lift coefficient and the bubble size. Particular attention is devoted to the implementation and extension of a clustering method ology for modeling the morphology-transfer, the Over-Resolved To Continuous (OTC) method, enabling the identification and conversion of over-resolved dispersed gas regions into continuous resolved bubbles, ensuring mass and momentum conservation during the transitions. The simulations demonstrate the capability of the MultiMorph framework to consistently represent both dispersed and resolved gas structures within the same com putational domain, maintaining numerical stability across mesh refinement levels. The proposed framework is first verified through a mesh convergence study and dedicated test cases, assessing numerical robustness and consistency across morphology transitions. Subsequently, the model is validated against experimental data from the literature, considering both monodispersed and polydispersed configurations and different mesh resolutions. Global quantities such as gas holdup and local features, including radial profiles of gas volume fraction, are analyzed and compared with experi mental measurements. The results show that the polydispersed formulation significantly improves the prediction of radial profiles. Results demonstrate that the morphology-adaptive approach successfully captures regime transitions, bubble clustering phenomena and the interactions between dispersed and resolved structures, while maintaining numerical stability. The integration of PBM further improves the predictive capability in polydispersed regimes. The study confirms the potential of the MultiMorph framework as a robust and scalable tool for advanced bubble column simulations and represents a foundational step toward the modeling of more complex industrial multiphase systems.| File | Dimensione | Formato | |
|---|---|---|---|
|
Thesis_Elena_Tagliabue.pdf
solo utenti autorizzati a partire dal 18/02/2029
Descrizione: Tesi di Laurea Magistrale in Chemical Engineering Elena Antonia Tagliabue
Dimensione
412.13 MB
Formato
Adobe PDF
|
412.13 MB | Adobe PDF | Visualizza/Apri |
|
Executive_Summary_Tagliabue.pdf
solo utenti autorizzati a partire dal 18/02/2029
Descrizione: Executive summary della tesi di Laurea Magistrale in Chemical Engineering Elena Antonia Tagliabue
Dimensione
87.48 MB
Formato
Adobe PDF
|
87.48 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/10589/250584