The overall purpose of this thesis work is the development and validation of Computational Fluid Dynamics (CFD) methods for the simulation of two-phase expansions, with a focus on organic fluids. In particular, the aim is to include the metastable phase during the calculation of thermodynamic properties for processes in which the non-equilibrium effects play a significant role. Two different models are tested: The homogeneous equilibrium model (HEM) and the homogeneous relaxation model (HRM). While the HEM assumes an instantaneous equilibrium between the vapor and liquid phase, the HRM takes into account the delay in the phase transition to capture metastability. In order to obtain accurate results in the simulations without increasing the computational load, a Look-Up Tables (LuT) approach is employed. The validation is carried out through comparisons with experimental measurements from two studies: the Moby Dick experiment, which investigates water flashing, and the Zhu & Elbel experiment, which focuses on R134a flashing in nozzles. This study points out the limitations of the HEM model and how HRM improves the simulation results. Furthermore, it can be observed that the direct application of the calibrated relaxation model for water does not simulate the flashing of organic fluid properly, hence indicating the need for particular tuning. For example, in the Zhu & Elbel experiment, the HEM predicts a mass flow rate with an error of 50–60%. By employing the HRM using the source terms calibrated for water, the error is significantly reduced, reaching values below 20%. It is only after calibrating the source term that the error falls below 10% in all three cases. The methodology is explained within this thesis and demonstrates how the calibrated relaxation model significantly enhances the reliability of CFD simulations to capture metastable phenomena in two-phase flow.
Lo scopo di questo lavoro di tesi è lo sviluppo e la validazione di metodi di fluidodinamica computazionale (CFD) per la simulazione di espansioni bifasiche con una particolare attenzione ai fluidi organici. In particolare, l’obiettivo è includere la fase metastabile durante il calcolo delle proprietà termodinamiche nei processi in cui gli effetti di non equilibrio svolgono un ruolo cruciale. Vengono testati due modelli diversi: il modello di equilibrio omogeneo (HEM) e il modello di rilassamento omogeneo (HRM). Mentre l’HEM assume un equilibrio istantaneo tra la fase liquida e quella di vapore, l’HRM tiene conto del ritardo nella transizione di fase per cogliere la metastabilità. Al fine di ottenere risultati accurati nelle simulazioni senza aumentare eccessivamente il carico computazionale, si impiega un approccio basato sulle Look-Up Tables (LuT). La validazione viene effettuata attraverso il confronto con le misurazioni sperimentali di due studi: l’esperimento Moby Dick, che analizza il flashing dell’acqua, e l’esperimento Zhu & Elbel, che esamina il flashing del fluido R134a. Questo studio mette in evidenza i limiti del modello HEM e come l’HRM migliori i risultati della simulazione. Inoltre, si osserva che l’applicazione diretta del modello di rilassamento calibrato per l’acqua, non risulta adeguata per simulare correttamente il flashing di fluidi organici, evidenziando quindi la necessità di una taratura specifica. Per esempio, nell’esperimento Zhu & Elbel, l’HEM predice la portata massica con un errore del 50/60%. Passando all’utilizzo dell’HRM con i termini di sorgente calibrati per acqua, l’errore è drasticamente ridotto, raggiungendo valori sotto il 20%. È solo dopo aver calibrato il termine sorgente che l’errore scende sotto il 10% in tutti e tre i casi. La metodologia è esposta in questa tesi e dimostra come tale calibrazione del modello di rilassamento incrementi sensibilmente l’affidabilità delle simulazioni CFD, consentendo di cogliere in modo più accurato i fenomeni metastabili nei flussi bifase.
Evaluation of a homogeneous relaxation model for flashing flows of organic fluids
Nessim, Andrea
2023/2024
Abstract
The overall purpose of this thesis work is the development and validation of Computational Fluid Dynamics (CFD) methods for the simulation of two-phase expansions, with a focus on organic fluids. In particular, the aim is to include the metastable phase during the calculation of thermodynamic properties for processes in which the non-equilibrium effects play a significant role. Two different models are tested: The homogeneous equilibrium model (HEM) and the homogeneous relaxation model (HRM). While the HEM assumes an instantaneous equilibrium between the vapor and liquid phase, the HRM takes into account the delay in the phase transition to capture metastability. In order to obtain accurate results in the simulations without increasing the computational load, a Look-Up Tables (LuT) approach is employed. The validation is carried out through comparisons with experimental measurements from two studies: the Moby Dick experiment, which investigates water flashing, and the Zhu & Elbel experiment, which focuses on R134a flashing in nozzles. This study points out the limitations of the HEM model and how HRM improves the simulation results. Furthermore, it can be observed that the direct application of the calibrated relaxation model for water does not simulate the flashing of organic fluid properly, hence indicating the need for particular tuning. For example, in the Zhu & Elbel experiment, the HEM predicts a mass flow rate with an error of 50–60%. By employing the HRM using the source terms calibrated for water, the error is significantly reduced, reaching values below 20%. It is only after calibrating the source term that the error falls below 10% in all three cases. The methodology is explained within this thesis and demonstrates how the calibrated relaxation model significantly enhances the reliability of CFD simulations to capture metastable phenomena in two-phase flow.File | Dimensione | Formato | |
---|---|---|---|
Evaluation of a Homogeneous Relaxation Model for Flashing Flows of Organic Fluids.pdf
solo utenti autorizzati a partire dal 03/03/2028
Dimensione
2.79 MB
Formato
Adobe PDF
|
2.79 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/10589/235358