In this work, the Direct Numerical Simulation of the Spray-A of the Engine Combustion Network (ECN) is performed by the use of the PARIS Simulator. In particular, the Lagrangian Point Particle (LPP) Module has been extensively investigated to study the atomization process. In the first chapters, this module has been studied: all the input parameters as well as the theory behind this technique have been analyzed. In order to save computational time of the thesis, an older Eulerian simulation of the Spray-A has been used to find the correct setting of the LPP module in order to tag and convert the highest possible number of droplets. Subsequently, a code written in Python called LPP_Finder has been developed. This code is able to open the result file provided by the Eulerian simulation and scan it to find all the produced liquid structures while at the same time providing a first statistical analysis of the results. Then a new Eulerian simulation of the Spray-A has been produced and subsequently two time steps using the LPP module conversion have been produced to tag all the possible droplets produced in during the simulation. It is going to be seen that the Python code and the LPP module’s results differ in both number of droplets tagged and droplets characteristics. While, both tool can capture the same droplets, the LPP_Finder captures and analyzes also all the small ligaments produced while the LPP module is unable to capture all the droplets produced due to the physical limitations of the theory describing the module. This produces a significant difference in the number of tagged liquid structures as well as in the distribution of the main values describing the droplets like the Sauter Mean Diameter (SMD). At the end, using the extra data provided by the LPP module, an analysis regarding the dimension, the position and the velocity of the tagged droplet is produced and commented. Since the LPP module can also be used in order to improve the simulation quality and save computational time, given the extensive experience gained in this work, some suggestions regarding other possible settings of this module have been provided.
In questa tesi, una “Direct Numerical Simulation” dello Spray-A dell’Engine Combustion Network (ECN) è stata simulata utilizzando il PARIS Simulator. Per studiare l’atomizzazione dello spray, viene proposto, nei primi capitoli, uno studio completo del modulo Lagrangiano (modulo LPP) analizzando tutti i parametri di input e la teoria alla base di questa tecnica. Per ridurre il tempo di calcolo di questa tesi, è stata utilizzata una vecchia simulazione Euleriana dello Spray-A per trovare il settaggio corretto del modulo LPP al fine di consentire l’individuazione del maggior numero possibile di gocce. Successivamente, è stato sviluppato un codice Python chiamato LPP_Finder. Questo codice è capace di aprire il file risultante dalla simulazione Euleriana e di analizzarlo per trovare e identificare tutte le strutture liquide prodotte durante la simulazione. Contemporaneamente il codice produce anche una prima analisi statistica dei risultati. Procedendo con il lavoro, su una nuova simulazione Euleriana dello Spray-A, sono stati prodotti 2 time-steps utilizzando il modulo LPP per permettere l’individuazione del maggior numero possibile di gocce prodotte. Analizzando i risultati prodotti dal modulo LPP e dall’LPP_Finder, si rileva che questi si differenziano sia nel numero di gocce identificate che nelle loro caratteristiche. Mentre entrambi questi strumenti trovano le stesse gocce, l’LPP_Finder identifica anche tutti i piccoli legamenti presenti mentre il modulo LPP non è in grado di identificare tutte le gocce prodotte a causa delle limitazioni fisiche del modulo stesso. Il risultato è una differenza significativa sia nel numero di strutture liquide trovate che nella distribuzione statistica delle loro caratteristiche. Inoltre è stata effettuata un’analisi statistica delle dimensioni, delle posizioni e delle velocità di tutte le gocce trovate dal modulo LPP sfruttando così tutti i dati prodotti dal PARIS Simulator. Infine, poiché il modulo LPP può essere utilizzato anche per ridurre il tempo di calcolo e migliorare la qualità della simulazione, vengono forniti alcuni suggerimenti addizionali utili per utilizzare questo modulo con altri obbiettivi.
Study of the atomization process of a direct numerical simulation of a Diesel spray by the use of the Lagrangian point particle method
PAOLETTI, MICHELE
2016/2017
Abstract
In this work, the Direct Numerical Simulation of the Spray-A of the Engine Combustion Network (ECN) is performed by the use of the PARIS Simulator. In particular, the Lagrangian Point Particle (LPP) Module has been extensively investigated to study the atomization process. In the first chapters, this module has been studied: all the input parameters as well as the theory behind this technique have been analyzed. In order to save computational time of the thesis, an older Eulerian simulation of the Spray-A has been used to find the correct setting of the LPP module in order to tag and convert the highest possible number of droplets. Subsequently, a code written in Python called LPP_Finder has been developed. This code is able to open the result file provided by the Eulerian simulation and scan it to find all the produced liquid structures while at the same time providing a first statistical analysis of the results. Then a new Eulerian simulation of the Spray-A has been produced and subsequently two time steps using the LPP module conversion have been produced to tag all the possible droplets produced in during the simulation. It is going to be seen that the Python code and the LPP module’s results differ in both number of droplets tagged and droplets characteristics. While, both tool can capture the same droplets, the LPP_Finder captures and analyzes also all the small ligaments produced while the LPP module is unable to capture all the droplets produced due to the physical limitations of the theory describing the module. This produces a significant difference in the number of tagged liquid structures as well as in the distribution of the main values describing the droplets like the Sauter Mean Diameter (SMD). At the end, using the extra data provided by the LPP module, an analysis regarding the dimension, the position and the velocity of the tagged droplet is produced and commented. Since the LPP module can also be used in order to improve the simulation quality and save computational time, given the extensive experience gained in this work, some suggestions regarding other possible settings of this module have been provided.| File | Dimensione | Formato | |
|---|---|---|---|
|
2017_07_PAOLETTI.pdf
accessibile in internet per tutti
Descrizione: Testo della Tesi
Dimensione
20.79 MB
Formato
Adobe PDF
|
20.79 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/10589/135467