The present work describes the processes of production, dissipation and transport of the Reynolds stress components in turbulent channel flow for three values of the Reynolds number, considering simultaneously the spaces of scales and positions. The tool used to perform the analysis is the Anisotropic Generalized Kolmogorov Equation (AGKE), a statistical budget equation for the second-order structure function hδu i δu j i, where δu i is the incre- ment of the i th velocity component at position X and separation r, i.e. δu i = (u i (X + r/2, t) − u i (X − r/2, t)). For the computation of the AGKE terms, three DNS-generated databases for turbulent channel flow at friction Reynolds number Re τ = 200, Re τ = 500 and Re τ = 1000 were considered. The analysis highlighted the presence of three peaks in the production of hδuδui. The first, common to all the three Reynolds numbers, involves the structures of the near-wall cycle. The second, observed for Re τ = 500 and Re τ = 1000 but not for Re τ = 200, was associated to the presence of attached eddies. The third, observed only for Re τ = 1000, was linked to the so-called large- and very large-scale motions. In the region of the first peak, near to the wall, the presence of transport processes of hδu i δu j i that do not scale in wall units highlighted the interaction with the outer scales. In the second peak, proportionality between the scale and the wall-normal distance of some trans- port processes and the inactivity at the wall of motions transporting hδvδvi were observed features. Moreover, an observed transfer of hδwδwi towards scales more and more spanwise-oriented for increasing wall-normal distances was associated to the presence of hairpin vortices in the flow. Finally, some features of this second peak were found to be visible also for Re τ = 200, suggesting a continuous transition from low- to high-Reynolds number flows. The third peak of production of hδuδui was also characterized, together with the associated transport of energy. The complete description provided by the AGKE in the compound space of scales and positions makes it an effective tool to study turbulent flows in presence of inhomogeneities. This description is even more important for high Reynolds numbers, for which the energy is distributed among a larger number of scales in different points of the physical space. In particular, it appears to be beneficial from both a theoretical and a modelling perspective. For example, it may be used to study flows to which drag reduction tech- niques have been applied, or to refine existing models used in large-eddies simulations, reproducing the effect of the small unresolved scales on the re- solved motion.
Il presente lavoro descrive i processi di produzione, dissipazione e trasporto delle componenti del tensore degli sforzi di Reynolds in correnti turbolente delimitate da pareti per tre valori del numero di Reynolds, considerando simultaneamente lo spazio delle scale e delle posizioni. Lo strumento utilizzato per effettuare l'analisi e' la Anisotropic Generalized Kolmogorov Equation (AGKE), un'equazione di bilancio statistica per la funzione di struttura del second'ordine <du_idu_j>, in cui du_i e' l'incremento della i-esima componente di velocita' nella posizione X e separazione r, i.e. du_i=(u_i(X + r/2,t) - u_i(X - r/2,t)). Tale equazione permette di cogliere i processi di trasporto di <du_idu_j> che avvengono simultaneamente tra fluttuazioni di diversa scala e diverse regioni della corrente. Per il calcolo dei termini dell'AGKE tre databases, prodotti mediante DNS per canale turbolento a numero di Reynolds viscoso pari a Re_\tau=200, Re_\tau=500 e Re_\tau=1000, sono stati considerati. Tale analisi ha evidenziato la presenza di tre picchi nella produzione di <dudu>. Il primo, comune a tutti e tre i numeri di Reynolds, coinvolge le strutture del ciclo di parete. Il secondo, osservato a Re_\tau=500 e Re_\tau=1000 ma non a Re_\tau=200, e' stato associato alla presenza di vortici attached. Il terzo, osservato solo per Re_\tau=1000, e' stato attribuito ai cosiddetti moti di grande e grandissima scala. Nella regione del primo picco, vicina a parete, la presenza di processi di trasporto di <du_idu_j> che non scalano in unita' viscose ha evidenziato l'interazione con le scale esterne. Nel secondo picco, la proporzionalita' tra la scala e la distanza da parete di alcuni processi di trasporto e l'inattivita' a parete dei moti che trasportano <dvdv> sono caratteristiche osservate. Inoltre, l'individuato trasferimento di <dwdw> verso scale con orientazione sempre piu' trasversale alla corrente all'aumentare della distanza da parete e' stato associato alla presenza di hairpin vortices. Infine, alcune caratteristiche di questo secondo picco sono state osservate anche per Re_\tau=200, supportando l'idea di una transizione continua da bassi ad alti numeri di Reynolds. Il terzo picco di produzione di <dudu> e' stato inoltre caratterizzato, insieme al trasporto di energia a questo associato. L'AGKE, offrendo una descrizione esaustiva nello spazio composto delle scale e delle posizioni, risulta costituire uno strumento efficace per lo studio di correnti turbolente in presenza di disomogeneita'. Tale descrizione acquista importanza ad elevati numeri di Reynolds, per i quali l'energia e' distribuita tra un maggiore numero di scale in diversi punti dello spazio fisico. In particolare, puo' avere un impatto rilevante sia da un punto di vista teorico sia per la modellazione. Per esempio, potrebbe essere impiegata per lo studio di correnti alle quali sono state applicate tecniche di drag reduction, o per affinare modelli esistenti utilizzati nelle simulazioni LES, riproducendo l'effetto delle piccole scale non risolte su quelle risolte.
Second-order structure function tensor budgets for channels at different values of Reynolds number
VILLANI, MARCO
2018/2019
Abstract
The present work describes the processes of production, dissipation and transport of the Reynolds stress components in turbulent channel flow for three values of the Reynolds number, considering simultaneously the spaces of scales and positions. The tool used to perform the analysis is the Anisotropic Generalized Kolmogorov Equation (AGKE), a statistical budget equation for the second-order structure function hδu i δu j i, where δu i is the incre- ment of the i th velocity component at position X and separation r, i.e. δu i = (u i (X + r/2, t) − u i (X − r/2, t)). For the computation of the AGKE terms, three DNS-generated databases for turbulent channel flow at friction Reynolds number Re τ = 200, Re τ = 500 and Re τ = 1000 were considered. The analysis highlighted the presence of three peaks in the production of hδuδui. The first, common to all the three Reynolds numbers, involves the structures of the near-wall cycle. The second, observed for Re τ = 500 and Re τ = 1000 but not for Re τ = 200, was associated to the presence of attached eddies. The third, observed only for Re τ = 1000, was linked to the so-called large- and very large-scale motions. In the region of the first peak, near to the wall, the presence of transport processes of hδu i δu j i that do not scale in wall units highlighted the interaction with the outer scales. In the second peak, proportionality between the scale and the wall-normal distance of some trans- port processes and the inactivity at the wall of motions transporting hδvδvi were observed features. Moreover, an observed transfer of hδwδwi towards scales more and more spanwise-oriented for increasing wall-normal distances was associated to the presence of hairpin vortices in the flow. Finally, some features of this second peak were found to be visible also for Re τ = 200, suggesting a continuous transition from low- to high-Reynolds number flows. The third peak of production of hδuδui was also characterized, together with the associated transport of energy. The complete description provided by the AGKE in the compound space of scales and positions makes it an effective tool to study turbulent flows in presence of inhomogeneities. This description is even more important for high Reynolds numbers, for which the energy is distributed among a larger number of scales in different points of the physical space. In particular, it appears to be beneficial from both a theoretical and a modelling perspective. For example, it may be used to study flows to which drag reduction tech- niques have been applied, or to refine existing models used in large-eddies simulations, reproducing the effect of the small unresolved scales on the re- solved motion.File | Dimensione | Formato | |
---|---|---|---|
Thesis_Villani.pdf
accessibile in internet per tutti
Descrizione: Testo della tesi
Dimensione
5.35 MB
Formato
Adobe PDF
|
5.35 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/10589/148407