The study of pressurized liquefied chlorine release is important in the context of environmental impact, emergency response, industrial safety and risk mitigation. In this study, a numerical model for the diffusion of chlorine leakage from a tank is implemented using the OpenFOAM software while it was used Paraview and Matlab for the postprocessing, visualization, and analysis of the results. The aim of this thesis is to create a simplified model that can meet the requirements of accuracy of chlorine concentration data and computational efficiency, since using the current models in full-scale problems would be computationally expensive due to their high accuracy. The validation of the results of the chlorine dispersion model is carried out by comparing the results obtained from the simulations with the experimental data from the Jack rabbit II test performed in a flat desert site within Dugway Proving Ground (DPG) in Utah between 2015 and 2016. Of the nine trials conducted during these experiments, Trial 3 was chosen as the reference experiment because it had the least variation in wind magnitude and direction. Initially, the scalarTransportFoam solver was used to achieve the objective, but given the same initial conditions, this solver fails to capture the variations that the wind undergoes over the time frame of the experiment. Therefore, the pimpleFoam solver was used with the function object scalarTransport, which solved the scalar transport equation simultaneously and captured the unsteadiness of the phenomenon. The parameter chosen for the comparison of results is the maximum concentration measured at specific arcs of distance from the source. The results of this model showed good agreement with data from the DRIFT model and the PHAST model, which are two of the main models used to reproduce these experiments, also accurately reproducing the concentration trends at distances of 200 meters and 500 meters. However, compared to the experimental data, the results obtained from this model are conservative and can be improved in more stable atmospheric conditions.
Lo studio del rilascio e della dispersione di cloro liquefatto pressurizzato è importante nel contesto dell’impatto ambientale, della risposta alle emergenze, della sicurezza industriale e della mitigazione dei rischi. In questo studio, viene implementato un modello numerico per la diffusione delle perdite di cloro da un serbatoio utilizzando il software OpenFOAM, mentre sono stati utilizzati Paraview e Matlab per il post-processing, la visualizzazione e l’analisi dei risultati. L’obiettivo di questa tesi è la creazione di un modello semplificato che possa soddisfare i requisiti di accuratezza dei dati sulla concentrazione del cloro e di efficienza computazionale, poiché l’utilizzo dei modelli attuali in problemi su scala reale sarebbe computazionalmente costoso a causa della loro elevata accuratezza. La validazione dei risultati del modello di dispersione del cloro viene effettuata confrontando i risultati ottenuti dalle simulazioni con i dati sperimentali del test Jack rabbit II eseguito in un sito desertico pianeggiante all’interno del Dugway Proving Ground (DPG) nello Utah tra il 2015 e il 2016. Delle nove prove condotte durante questi esperimenti, il Trial 3 è stato scelto come esperimento di riferimento perché presentava la minore variazione di ampiezza e direzione del vento. Inizialmente, per raggiungere l’obiettivo è stato utilizzato il solver scalarT ransportF oam, ma a parità di condizioni iniziali, questo solutore non riesce a catturare le variazioni che il vento subisce nell’arco temporale dell’esperimento. Pertanto, è stato utilizzato il solver pimpleF oam con la function object scalarT ransport, che risolve l’equazione di trasporto scalare simultaneamente e coglie l’instabilità del fenomeno. Il parametro scelto per la comparazione dei risultati è la concentrazione massima misurata a specifici archi di distanza dalla sorgente. I risultati di questo modello hanno mostrato un buon accordo con i dati del modello DRIFT e del modello PHAST, che sono due dei principali modelli utilizzati per riprodurre questi esperimenti, riproducendo accuratamente anche l’andamento della concentrazione a distanze di 200 e 500 metri. Tuttavia, rispetto ai dati sperimentali, i risultati ottenuti da questo modello sono conservativi e possono essere migliorati in condizioni atmosferiche più stabili.
CFD analysis of chlorine dispersion in an urban area: a comparative study with Jack Rabbit II test data
D'Ignazio, Maurizio
2022/2023
Abstract
The study of pressurized liquefied chlorine release is important in the context of environmental impact, emergency response, industrial safety and risk mitigation. In this study, a numerical model for the diffusion of chlorine leakage from a tank is implemented using the OpenFOAM software while it was used Paraview and Matlab for the postprocessing, visualization, and analysis of the results. The aim of this thesis is to create a simplified model that can meet the requirements of accuracy of chlorine concentration data and computational efficiency, since using the current models in full-scale problems would be computationally expensive due to their high accuracy. The validation of the results of the chlorine dispersion model is carried out by comparing the results obtained from the simulations with the experimental data from the Jack rabbit II test performed in a flat desert site within Dugway Proving Ground (DPG) in Utah between 2015 and 2016. Of the nine trials conducted during these experiments, Trial 3 was chosen as the reference experiment because it had the least variation in wind magnitude and direction. Initially, the scalarTransportFoam solver was used to achieve the objective, but given the same initial conditions, this solver fails to capture the variations that the wind undergoes over the time frame of the experiment. Therefore, the pimpleFoam solver was used with the function object scalarTransport, which solved the scalar transport equation simultaneously and captured the unsteadiness of the phenomenon. The parameter chosen for the comparison of results is the maximum concentration measured at specific arcs of distance from the source. The results of this model showed good agreement with data from the DRIFT model and the PHAST model, which are two of the main models used to reproduce these experiments, also accurately reproducing the concentration trends at distances of 200 meters and 500 meters. However, compared to the experimental data, the results obtained from this model are conservative and can be improved in more stable atmospheric conditions.File | Dimensione | Formato | |
---|---|---|---|
2024_04_D'IGNAZIO.pdf
solo utenti autorizzati a partire dal 19/03/2027
Descrizione: Testo tesi
Dimensione
12.88 MB
Formato
Adobe PDF
|
12.88 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/10589/218520