9,075 research outputs found
Backward energy transfer and subgrid modeling approaches in wall-turbulence
No full textWe report here results from a Large Eddy Simulation (LES) of a turbulent channel flow at a friction Reynolds number Reτ = 550 performed with a new subgrid modeling approach proposed by the same authors in Cimarelli et al., Phys. Fluids, 26, 055103 (2014), [1]. This subgrid scale model aims at reproducing the double feature of energy sink and source of the small scales of wall flows which become relevant when large filter lengths are adopted. Here we report a further analysis of the model by considering the instantaneous behavior of events of backward and forward energy transfer
Droplet condensation in turbulent jets
No full textThe evaporation and condensation process of liquid droplets, advected by a turbu- lent flow is found in many technological applications - ranging from aeronautical/power- plants engines up to the rather new discipline of Particle Engineering in the biotech industry - and in many natural phenomena, such as cloud formation and meteorology. A better understanding of droplets nucleation, transport, evaporation/condensation and vapour mixing is crucial for the efficiency of these devices as well as it would be to improve weather forecasting.
The phase change is a multi-scale phenomena ranging from the nano scale, at which droplets nucleate, to the micro and macro scales where the turbulent flow advects both the droplets and their vapour. Although droplet laden flows have been extensively studied, several issues are still in place, especially when dealing with multi-phase turbulent flows. Many phenomena, such as small scales clustering of droplets (considered as inertial particles) or preferential spatial accumulation, have been observed and understood under the simplifying assumption of the one-way cou- pling regime, where the disperse phase does not modify the carrier fluid. Nonetheless, a deeper understanding of multiphase flows demands at least for two-way coupling effects, accounting for the inter-phase mass, momentum and energy exchanges. Homogeneous nucleation of liquid droplets occurs when a hot vapour stream mixes with a cooler and dry external flow. Even though many applications could benefit from a better understanding of droplet nucleation in turbulence, nowadays it has been investigated only experimentally, given all the aforementioned difficulties on modelling the multi-physics phenomena occurring in such a complex turbulent flow. In fact, the nonlinear interplay between turbulent fluctuations and homogeneous nucleation, immediately leads to non-trivial cross-coupling phenomena between the gas and the liquid phase. Their modelling reveals to be crucial for detailed numerical investigations to be reliable.
Classical Nucleation Theory (CNT) prescribes a rate for the homogenous nucleation of droplets, per unit time and volume. It also provides for an estimate of the critical radius at which – eventually – each droplet nucleates. The droplet nucleation itself basically occurs when a stable molecules cluster size is reached, thus everything strongly depends on the local thermodynamical state. While the equations describ- ing mass, momentum and energy exchange, between droplets and the turbulent flow are still formulated on phenomenological ground, a rigorous derivation of the fluid flow equations still lacks. In the present work a first attempt to tackle this challenging physical problem will be presented: modelling the mass, momentum and energy transfer, with appropriate boundary conditions, at the interface between the two phases. Starting from the low-Mach number formulation of the Navier-Stokes equations for the gaseous phase, an analytical decomposition of the flow field allows
– within the point particle approximation, for the disperse phase – to reallocate the boundary conditions at the droplets surface as equivalent source/sink terms on the carrier phase, without any ad-hoc assumption. However this methodology still needs an estimate for the fluxes at the droplet surface, at any length-scale. Two different models for the mass transfer, have been tested and employed here to be validated against the available experimental investigations found in literature.
In the present study a full description of the droplet nucleation in a vapour turbulent jet is provided by means of Direct Numerical Simulation (DNS). No turbulence model has been considered, since all the relevant scales of turbulent motion have been resolved on the computational grid. To capture the inter-phase exchanges the so-called Exact Regularised Point-Particle method (ERPP) has been adopted, proving again its suitability for High Performance Computing simulations (HPC) on massively parallel machines, handling billions particles. In the point-particle ap- proximation and within the Eulerian-Lagrangian approach, here adopted to describe respectively the carrier and disperse phase dynamics, each droplets is described as point mass. The relevance of the inter-phase coupling effects is thoroughly discussed, by comparing the results obtained accounting for the droplets back-reaction with those obtained neglecting it. It is already known how a disperse phase does affect the carrier phase dynamics, modulating turbulence, stretching the shape of the jet, but always preserving the (statuary) jet self-similarity found in the one-way coupling regime. Beyond turbulence modulation, the droplets back-reaction is mainly effective on the temperature and vapour fields of the carrier phase, heavily altering the local thermodynamical fluctuations, thus affecting the related droplet nucleation rate. This way, one is able to completely characterise the phase change process, to measure any droplet trajectory followed in a Lagrangian way and to fully characterise any observable from a statistical point of view.
Considerable effects of the droplets back-reaction on the nucleation rate emerged, by decreasing the amount of vapour – due to nucleation/condensation – and heating up, to a lesser extent, the dry environment. Both effects decrease the vapour saturation and the nucleation rate that strongly depends on it. The intensity of these effects is related to the number of back-reacting nucleated droplets, thus at lower concen- trations the nucleation rate is almost insensitive to it. However, for higher vapour concentration, it is not possible to neglect the particles back-reaction, especially being these an additional source of turbulent fluctuations that ultimately impacts on the mean nucleation rate. In fact, the droplets redistribute vapour through condensation/evaporation and temperature by releasing/absorbing heat, along their trajectory. In other words, the particles back-reaction serves as an additional source of fluctuations not to be disregarded. It is reasonable to hypothesise that it is not possible to effectively model the back-reaction effects, discussed so far, rather than to simply evolve each single droplet as it has been done in the present DNS simulations. Under this respect, the present contribution constitutes an absolute novelty in the aerosol community shedding light on the the two-way coupling effects that were revealed to be crucial in the overall nucleation process and turbulent transport of the droplets
Routes to chaos of natural convection flows in vertical channels
No full textThe aim of the present study is the analysis of the transition to turbulence of natural convection flows between two infinite vertical plates. For the study of the problem, a number of Direct Numerical Simulations (DNSs) have been performed. The continuity, momentum and energy equations, cast under the Boussinesq assumption, are tackled numerically by means of a pseudospectral method, through which the three-dimensional domain is decomposed with Chebychev polynomials in the wall-normal direction and with Fourier modes in the wall-parallel directions. For low Rayleigh number values, the predictions of the flow regimes are consistent with the classical analytical results and linear stability analyses. In particular, the first bifurcation (Ra ≈ 5800) from the so-called laminar conduction regime to steady convection is correctly captured. By increasing the Rayleigh number beyond a second critical value (Ra ≈ 10200), the flow regime becomes chaotic. This transition to chaos is found to be related with the amplification of spanwise instabilities occurring at scales larger than the channel gap, H. The study of the return of the system from the chaotic regime to the laminar base flow reveals a phenomenon of hysteresis, i.e. the chaotic regime persists even at Ra-values lower than the second critical value. From a numerical point of view, the predicted flow regimes appear to be extremely sensitive to the domain size, grid resolution and perturbation amplitude. These aspects are shown to be of crucial importance for the prediction of the heat transfer performance, and, hence, should be taken into consideration when numerical methods are used for the simulation of real-world problems
Turbulent production and subgrid dynamics in wall flows
No full textThe Kolmogorov equation generalized to wall-turbulence has been recently proven to give a detailed description of the multi-dimensional features of such flows[1]. As emerging from this approach, the small scales of wall turbulence are found to drive the quasi-coherent motion at large scales through a reverse energy transfer. At the base of this phenomenology is the focusing of production of turbulent fluctuations at small scales. These observations may have strong repercussion on both theoretical and modeling approaches to wall-turbulence. Here, we aim at using the Kolmogorov equation not only for the study of the mechanisms altering the energy transfer but also for modeling purpose
Direct numerical simulation of the flow around a rectangular cylinder at a moderately high Reynolds number
No full textWe report a Direct Numerical Simulation (DNS) of the flow around a rectangular cylinder with a chord-to-thickness ratio B/D=5 and Reynolds number Re=3000. Global and single-point statistics are analysed with particular attention to those relevant for industrial applications such as the behaviour of the mean pressure coefficient and of its variance. The mean and turbulent flow is also assessed. Three main recirculating regions are found and their dimensions and turbulence levels are characterized. The analysis extends also to the asymptotic recovery of the equilibrium conditions for self-similarity in the fully developed wake. Finally, by means of two-point statistics, the main unsteadinesses and the strong anisotropy of the flow are highlighted. The overall aim is to shed light on the main physical mechanisms driving the complex behaviour of separating and reattaching flows. Furthermore, we provide well-converged statistics not affected by turbulence modelling and mesh resolution issues. Hence, the present results can also be used to quantify the influence of numerical and modelling inaccuracies on relevant statistics for the applications
Assessment of the turbulent energy paths from the origin to dissipation in wall-turbulence
No full textThe present study is devoted to the description of the energy fluxes from production to dissipation in the augmented space (3-dimensional space of scales plus wall-distance) of wall-turbulent flows. As already shown in Cimarelli et al. (2010), an interesting behavior of the energy fluxes comes out from this analysis consisting of spiral-like paths in the combined physical/scale space where the controversial reverse energy cascade plays a central role. The observed behaviour conflicts with the classical notion of the Richardson/Kolmogorov energy cascade and may have strong repercussions on both theoretical and modeling approaches to wall-turbulence. Two dynamical processes are identified as driving mechanisms for the fluxes, one in the near wall region and a second one further away from the wall. The former, stronger one is related to the dynamics involved in the near-wall cycle. The second suggests an outer self-sustaining mechanism. Here we extend these results to larger Reynolds number using LES data of a turbulent channel flow at Re τ = 970 confirming the presence of an outer regeneration cycle which seems to be composed by systems of attached eddies
Physical and scale-by-scale analysis of Rayleigh-Bénard convection
No full textA novel approach for the study of turbulent Rayleigh-Bénard convection (RBC) in the compound physical/scale space domain is presented. All data come from direct numerical simulations of turbulent RBC in a laterally unbounded domain confined between two horizontal walls, for Prandtl number 0:7 and Rayleigh numbers 1:7 ± 105, 1:0 ± 106 and 1:0 ± 107. A preliminary analysis of the flow topology focuses on the events of impingement and emission of thermal plumes, which are identified here in terms of the horizontal divergence of the instantaneous velocity field. The flow dynamics is then described in more detail in terms of turbulent kinetic energy and temperature variance budgets. Three distinct regions where turbulent fluctuations are produced, transferred and finally dissipated are identified: a bulk region, a transitional layer and a boundary layer. A description of turbulent RBC dynamics in both physical and scale space is finally presented, completing the classic single-point balances. Detailed scale-by-scale budgets for the second-order velocity and temperature structure functions are shown for different geometrical locations. An unexpected behaviour is observed in both the viscous and thermal transitional layers consisting of a diffusive reverse transfer from small to large scales of velocity and temperature fluctuations. Through the analysis of the instantaneous field in terms of the horizontal divergence, it is found that the enlargement of thermal plumes following the impingement represents the triggering mechanism which entails the reverse transfer. The coupling of this reverse transfer with the spatial transport towards the wall is an interesting mechanism found at the basis of some peculiar aspects of the flow. As an example, it is found that, during the impingement, the presence of the wall is felt by the plumes through the pressure field mainly at large scales. These and other peculiar aspects shed light on the role of thermal plumes in the self-sustained cycle of turbulence in RBC, and may have strong repercussions on both theoretical and modelling approaches to convective turbulence
L"interazione radiazione-materia
No full textCD divulgativo progettato per le scuole medie superiori, che introduce al tema delle interazioni radiazione-materia e alle loro applicazioni in campo chimico
Numerical Experiments on Turbulent Entrainment
No full textThe aim of this thesis work is the study of the turbulent entrainment phenomenon in jets through numerical experiments. More specifically, an attempt to study the effect of engulfment and nibbling mechanisms separately was made. The flow chosen for the numerical experiments is the temporal planar jet. The idea behind these experiments is to study the spreading and the mixing of a passive scalar under the effect of two modified velocity fields. The first is a large-scale velocity field obtained through a filtering operation, while the second is a small-scale velocity field obtained subtracting the large-scale velocity field from the total one and then adding the mean velocity.
Initially, the post-processing of a spatially developing planar jet, performed by Doctor Andrea Fregni and Professor Andrea Cimarelli, has been carried out in order to analyse the main features of spatially evolving jets compared with the temporal ones. A co-flow and a passive scalar are present in the simulation. The Reynolds number is set to Re = 3000 and the Schmidt number is Sc = 1. After this first step, a benchmark DNS of a temporal planar jet with Re = 3000 and Sc = 1 has been performed in order to evaluate the main differences with respect to the spatially evolving jet. Once the settings were validated, the numerical experiments with large and small scale velocity fields have been performed. The filter used in all the experiments is the box filter. The results of two different filter lengths are presented, the first is Δ = 1.5λcl and the second is Δ = 3λcl. Since λcl is function of time, the two filter lengths are themselves varying in time. The results of the experiments were then compared with those of the unfiltered solution. The passive scalar spread approximatively the same amount under the effect of the large-scale velocity fields and under the effect of the unfiltered velocity. On the other hand, the small-scale fluctuations have been proved to be important in the mixing process
Direct simulation of transition in a differentially heated vertical channel
No full textTransition to turbulence of natural convection flows ensuing in a fluid layer between two differentially heated vertical plates is a topic of substantial interest for many applications. Among these, notable examples are the air gaps in double-glazing panes or in ventilated façades, and passive heat exchangers. The correct prediction and control of flow regimes, air flow rates and heat transfer coefficients has a significant impact in the correct design of such elements and, in turn, on their efficiency.
In recent studies the early stages of transition have been explored by means of Direct Numerical Simulation (DNS) with high-accuracy pseudospectral codes. While all these studies correctly capture the first bifurcation from the so-called laminar conduction regime to steady convection, the detection of the subsequent transition to turbulence appears to be accompanied by a great sensitivity to some fundamental numerical choices, such as domain size, spectral resolution and amplitude of the imposed perturbations. In turn, these aspects become of crucial importance for the prediction of the heat transfer performance of the system.
In this work, the problem is tackled by means of a second-order, Finite-Volume based Direct Numerical Simulation technique, specifically devised for convection problems, and which already proved successful in the simulation of transitional scenarios. Results reveal the occurrence of a bifurcation branch which leads the system to chaos via a second bifurcation to a steady-state, a Hopf bifurcation and, seemingly, a period-doubling cascade. Such a scenario compares well with previous findings, except for minor discrepancies. All in all, though, some doubts persist upon the possible pitfalls in the use of DNS for the study of transition in this kind of systems
- …
