|Tipo di tesi||Tesi di laurea magistrale|
|Autore||DELLE CURTI, ANTONIO|
|Titolo||Sviluppo di uno strumento numerico adatto all'analisi aerodinamica e aeroacustica per autovetture|
|Titolo in inglese||Development of a numerical tool suitable for aerodynamic and aeroacoustic analysis of passenger vehicles|
|Struttura||Dipartimento di Ingegneria "Enzo Ferrari"|
|Corso di studi||Ingegneria Del Veicolo (D.M.270/04)|
|Data inizio appello||2022-07-14|
|Disponibilità||Accesso limitato: si può decidere quali file della tesi rendere accessibili. Disponibilità mixed (scegli questa opzione se vuoi rendere inaccessibili tutti i file della tesi o parte di essi)|
|Data di rilascio||2062-07-14|
Nel presente lavoro vengono presentati diversi strumenti numerici progettati per effettuare analisi aerodinamiche e aeroacustiche di veicoli passeggeri. Il comportamento aerodinamico di un modello di DrivAer in scala ridotta e di uno in scala reale è stato studiato attraverso la simulazione di Navier-Stokes mediata da Reynolds (RANS) e la simulazione DDES (Delayed-Detached-Eddy). Per l'intero studio è stato utilizzato il software open-source OpenFOAM. I modelli di turbolenza kω-SST e Spalart-Allmaras sono stati utilizzati rispettivamente per i modelli numerici stazionari (RANS) e non stazionari (DDES).
In the present work, different numerical tools designed to carry out aerodynamic and aeroacoustic analysis for passenger vehicles are presented. The aerodynamic behavior of a quarter-scale and a full-scale DrivAer model is investigated by means of Reynolds-averaged Navier-Stokes (RANS) and Delayed-Detached-Eddy Simulation (DDES) . The open-source software OpenFOAM has been used for the entire study. The turbulence models kω-SST and Spalart-Allmaras have been used for the steady (RANS) and the unsteady (DDES) numerical models, respectively . In the first part of the study, the CFD results concerning the DrivAer quarter-scale model have been validated against the experimental data taken in the model scale wind tunnel (1:4) of the university of Stuttgart operated by FKFS. Both simulations and measurements have been performed considering a freestream velocity of 45 m/s, and thus a Reynolds number of 3.2x10^6. The change of the aerodynamic coefficients as well as the velocity field with respect to a geometry variation have been investigated. The notchback and the estateback configurations with a detailed underbody have been used for this analysis. In the second part of the study, a DrivAer full-scale configuration has been investigated. The numerical predictions have been validated against experiments carried out in the Pininfarina full scale wind tunnel. The simulations as well as the related measurements have been performed considering a freestream velocity of 38.88 m/s, and thus a Reynolds number of 7.2x10^6.The validation based on the analysis of the aerodynamic coefficient and the velocity field is discussed. Both studies have been carried out with a static wheel model and a closed cooling system. Convincing correlations between CFD predictions and wind tunnel measurements have been found for both the cases investigated. RANS have shown a good level of accuracy in terms of aerodynamic coefficient variation with respect to a geometry change as well as in the velocity field modelling. DDES have been proven to improve the accuracy of the predictions although a significant increase of the computational cost has been observed. The predicted drag coefficient has shown deviations from experiments lower than -7 and 15 counts for the quarter-scale and the full-scale model, respectively. However, larger deviations have been observed regarding the lift coefficient for both RANS and DDES in the two studies considered. Furthermore, the predicted velocity field has been found to be in good agreement with experiments: percentage errors lower than 15% have been observed in both the studies investigated.