In the pursuit of cleaner and more efficient internal combustion engines, dual-fuel strategies combining diesel and hydrogen are gaining increasing attention for heavy duty applications. This thesis aims to carry out an experimental validation of the data provided by the University of New South Wales (UNSW) through the analysis and modeling of the combustion phenomena occurring inside a Hydrogen–Diesel Dual Direct Injection (H2DDI) engine, using Computational Fluid Dynamics (CFD) simulations. The numerical framework utilizes a Reynolds-Averaged Navier-Stokes (RANS) approach coupled with the Tabulated Well-Mixed (TWM) combustion model. The computational setup is extensively validated against experimental in-cylinder pressure measurements and high-speed natural flame luminosity imaging acquired from a large-bore optical engine. The research methodology initially focuses on the isolated calibration of the diesel pilot injection. Subsequently, the investigation proceeds with the calibration of the dual-fuel combustion model. This phase establishes a detailed numerical setup characterised by a sequential injection strategy, wherein the diesel pilot injection precedes the main hydrogen jet, alongside a baseline hydrogen energy share of 90%. To elucidate the complex thermochemical interactions between the two fuels and to investigate the morphological discrepancies observed between numerical predictions and experimental imaging, a comprehensive diagnostic analysis was conducted. The calibrated setup is further validated against two distinct experimental datasets, encompassing wide variations in both the Start of Injection (SOI) timings and the hydrogen-diesel energy proportions. Subsequent dual-fuel investigations demonstrate that the TWM model accurately reproduces global thermodynamic parameters and the steady-state mixing-controlled diffusion combustion phase. However, discrepancies emerged during the early ignition transient. Due to its intrinsic assumption of sub-grid homogeneity, the TWM formulation systematically under-predicts the initial premixed hydrogen heat release peak, as it fails to resolve the local mixture stratifications driving the auto-ignition process. Despite these limitations in capturing micro-scale ignition physics, the proposed RANS-TWM framework proves exceptionally computationally efficient, providing a highly robust tool for macroscopic parametric screening and thermodynamic cycle optimization.
L’integrazione dell’idrogeno nei motori diesel per mezzi pesanti in configurazione bi-combustibile emerge oggi come una delle soluzioni tecnologiche più promettenti per migliorare l’efficienza termodinamica e ridurre l’impatto ambientale. Il presente elaborato di tesi si propone di condurre una validazione sperimentale dei dati messi a disposizione dalla University of New South Wales (UNSW), mediante l’analisi e la modellazione dei fenomeni di combustione che si sviluppano all’interno di un motore a doppia iniezione diretta di idrogeno e diesel (H2DDI), avvalendosi di simulazioni Computational Fluid Dynamics (CFD). Dal punto di vista numerico la turbolenza è stata modellata mediante un approccio Reynolds-Averaged Navier-Stokes (RANS), mentre la chimica della combustione è stata descritta tramite il modello Tabulated Well-Mixed (TWM). Inizialmente, l’attenzione si è concentrata sulla calibrazione dell’iniezione pilota di diesel. La trattazione si è poi estesa alla combustione bicombustibile, riproducendo una strategia di iniezione nella quale il diesel viene iniettato per primo con funzione di innesco, mentre l’idrogeno viene immesso successivamente e costituisce il 90% dell’apporto energetico totale. Per verificare l’affidabilità del modello, la configurazione numerica è stata confrontata con le immagini ottiche ritraenti l’interno della camera di combustione. Un’indagine diagnostica ha permesso di comprendere a fondo le interazioni termochimiche tra i due combustibili e di spiegare le differenze morfologiche emerse tra la simulazione e le immagini provenienti dal motore ottico. Infine, la configurazione scelta è stata testata su due insiemi di dati sperimentali, variando in un caso il momento di apertura dell’iniettore diesel, nell’altro la proporzione energetica tra diesel e idrogeno. I risultati dimostrano che il modello TWM riproduce con precisione l’andamento termodinamico globale del motore e l’intera fase di combustione diffusiva, ma emergono alcune discrepanze durante i primi istanti di accensione dell’idrogeno. Questo accade poiché il modello TWM non è in grado di cogliere le stratificazioni locali di concentrazione e temperatura, in quanto ipotizza che la miscela sia perfettamente omogenea all’interno di ogni cella. Da questo limite intrinseco deriva una sottostima del picco iniziale di rilascio del calore associato alla fase di combustione premiscelata dell’idrogeno. Nonostante ciò, l’approccio RANS-TWM rappresenta un ottimo strumento per ottimizzare il ciclo termodinamico del motore su scala macroscopica.
CFD modeling of hydrogen-diesel dual direct injection for heavy-duty applications
DEIDDA, GIULIO;Casu, Giovanni
2024/2025
Abstract
In the pursuit of cleaner and more efficient internal combustion engines, dual-fuel strategies combining diesel and hydrogen are gaining increasing attention for heavy duty applications. This thesis aims to carry out an experimental validation of the data provided by the University of New South Wales (UNSW) through the analysis and modeling of the combustion phenomena occurring inside a Hydrogen–Diesel Dual Direct Injection (H2DDI) engine, using Computational Fluid Dynamics (CFD) simulations. The numerical framework utilizes a Reynolds-Averaged Navier-Stokes (RANS) approach coupled with the Tabulated Well-Mixed (TWM) combustion model. The computational setup is extensively validated against experimental in-cylinder pressure measurements and high-speed natural flame luminosity imaging acquired from a large-bore optical engine. The research methodology initially focuses on the isolated calibration of the diesel pilot injection. Subsequently, the investigation proceeds with the calibration of the dual-fuel combustion model. This phase establishes a detailed numerical setup characterised by a sequential injection strategy, wherein the diesel pilot injection precedes the main hydrogen jet, alongside a baseline hydrogen energy share of 90%. To elucidate the complex thermochemical interactions between the two fuels and to investigate the morphological discrepancies observed between numerical predictions and experimental imaging, a comprehensive diagnostic analysis was conducted. The calibrated setup is further validated against two distinct experimental datasets, encompassing wide variations in both the Start of Injection (SOI) timings and the hydrogen-diesel energy proportions. Subsequent dual-fuel investigations demonstrate that the TWM model accurately reproduces global thermodynamic parameters and the steady-state mixing-controlled diffusion combustion phase. However, discrepancies emerged during the early ignition transient. Due to its intrinsic assumption of sub-grid homogeneity, the TWM formulation systematically under-predicts the initial premixed hydrogen heat release peak, as it fails to resolve the local mixture stratifications driving the auto-ignition process. Despite these limitations in capturing micro-scale ignition physics, the proposed RANS-TWM framework proves exceptionally computationally efficient, providing a highly robust tool for macroscopic parametric screening and thermodynamic cycle optimization.| File | Dimensione | Formato | |
|---|---|---|---|
|
CFD_Modeling_of_Hydrogen_Diesel_Dual_Direct_Injection_for_Heavy_Duty_Applications.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Thesis
Dimensione
46.01 MB
Formato
Adobe PDF
|
46.01 MB | Adobe PDF | Visualizza/Apri |
|
Executive_Summary___CFD_Modeling_of_Hydrogen_Diesel_Dual_Direct_Injection_for_Heavy__Duty_Applications.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Executive Summary
Dimensione
7.37 MB
Formato
Adobe PDF
|
7.37 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/10589/251621