In light of recent developements in the field of turbulence control, the current work considers the link between the friction Reynolds number, hence the drag-reducing control, and the flow regime for the pipe flow. To do so, it seeks to identify a indicator, able to discern the regime of a pipe flow subjected to turbulence control in the form of spanwise forcing. The first query was assessed by comparing the state of two flows forced with different control parameters, but characterized by the same Reτ, which ultimately showed two very distinct flow conditions. The search for an indicator was achieved by applying three different methods, based on the analysis of the flow's statistics. Results of the three strategies were then compared. The algorithms, first validated on three selected flows, representatives of the laminar, fully turbulent and transitional regime, were then applied to a larger pool of sample flows, forced by waves with constant amplitude, but different phase speeds. All simulations were performed exploiting a pseudo-spectral, parallel DNS code. VISA results have proved to be the most satisfying, as they succed in separating the flow regimes in the kx − ω plane, based on the length of the events detected. As expected, no events are registered in case of laminar flow. The ratio between the average event length for the locally turbulent and the fully turbulent flows is discretely large, with lower values obtained when the turbulent flow is affected by DI. Further supported by the comparison with the drag reduction for spanwise forcing map, the method can be considered effective and very easy to implement. Only major downside is represented by the dependence of the results on the analysis parameters and especially on the averaging interval wx, flaws shared with all methods employed in the analysis.
Alla luce dei recenti sviluppi nel campo del controllo della turbolenza, il presente lavoro intende indagare la presenza di un eventuale legame tra il numero di Reynolds d'attrito, legato all'effetto di riduzione di resistenza del controllo, e il regime del flusso all'interno di un condotto cilindrico. Per fare ciò, si propone di identificare un indicatore in grado di discernere il regime di un flusso soggetto a forzamento trasversale per ridurne la resistenza d'attrito. La prima indagine ha subito riscontrato regimi diversi, seppur in presenza di stesso Reτ. Il passaggio successivo si è svolto applicando tre diversi metodi, basati sull'analisi delle statistiche del flusso. I risultati delle tre strategie sono stati poi confrontati. Gli algoritmi, prima validati su tre flussi selezionati, rappresentativi del regime laminare, turbolento e di transizione, sono stati poi applicati a un ampio database di ' flussi campione', forzati da onde con ampiezza costante ma diverse velocità di fase. Tutte le simulazioni sono state eseguite sfruttando un codice parallelo, pseudo-spettrale per la DNS di un flusso in condotto cilindrico. I risultati del metodo VISA si sono rivelati essere i più soddisfacenti, poiché riescono a distinguere i regimi di flusso nel piano kx − ω in base alla lunghezza degli eventi rilevati. Come previsto, non si registrano eventi nel caso di flusso laminare. Il rapporto tra la lunghezza media degli eventi per i flussi localmente turbolenti e quelli completamente turbolenti è discretamente ampio, con valori inferiori riscontrati quando il flusso turbolento è affetto da DI. Ulteriormente supportato dal confronto con i risultati della mappa di riduzione di resistenza ottenuta a seguito dell'applicazione di forzamento trasversale, il metodo può essere considerato efficace e di facile implementazione. L'unico principale svantaggio è rappresentato dalla dipendenza dei risultati dai parametri di analisi e in particolare dall'intervallo di media w_x, difetti, in ogni caso, condivisi con tutti i metodi impiegati nell'analisi.
Transition to turbulence in drag-reduced pipe flows
Ghilardi, Fabiana
2023/2024
Abstract
In light of recent developements in the field of turbulence control, the current work considers the link between the friction Reynolds number, hence the drag-reducing control, and the flow regime for the pipe flow. To do so, it seeks to identify a indicator, able to discern the regime of a pipe flow subjected to turbulence control in the form of spanwise forcing. The first query was assessed by comparing the state of two flows forced with different control parameters, but characterized by the same Reτ, which ultimately showed two very distinct flow conditions. The search for an indicator was achieved by applying three different methods, based on the analysis of the flow's statistics. Results of the three strategies were then compared. The algorithms, first validated on three selected flows, representatives of the laminar, fully turbulent and transitional regime, were then applied to a larger pool of sample flows, forced by waves with constant amplitude, but different phase speeds. All simulations were performed exploiting a pseudo-spectral, parallel DNS code. VISA results have proved to be the most satisfying, as they succed in separating the flow regimes in the kx − ω plane, based on the length of the events detected. As expected, no events are registered in case of laminar flow. The ratio between the average event length for the locally turbulent and the fully turbulent flows is discretely large, with lower values obtained when the turbulent flow is affected by DI. Further supported by the comparison with the drag reduction for spanwise forcing map, the method can be considered effective and very easy to implement. Only major downside is represented by the dependence of the results on the analysis parameters and especially on the averaging interval wx, flaws shared with all methods employed in the analysis.File | Dimensione | Formato | |
---|---|---|---|
2024_12_Ghilardi.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: testo tesi
Dimensione
7.06 MB
Formato
Adobe PDF
|
7.06 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/10589/230863