36 resultados para Turbulent Mixing
Resumo:
Numerical study has been performed in this study to investigate the turbulent convection heat transfer on a rectangular plate mounted over a flat surface. Thermal and fluid dynamic performances of extended surfaces having various types of lateral perforations with square, circular, triangular and hexagonal cross sections are investigated. RANS (Reynolds averaged Navier–Stokes) based modified k–ω turbulence model is used to calculate the fluid flow and heat transfer parameters. Numerical results are compared with the results of previously published experimental data and obtained results are in reasonable agreement. Flow and heat transfer parameters are presented for Reynolds numbers from 2000 to 5000 based on the fin thickness.
Resumo:
In an estuary, mixing and dispersion result from a combination of large-scale advection and smallscale turbulence, which are complex to estimate. The predictions of scalar transport and mixing are often inferred and rarely accurate, due to inadequate understanding of the contributions of these difference scales to estuarine recirculation. A multi-device field study was conducted in a small sub-tropical estuary under neap tide conditions with near-zero fresh water discharge for about 48 hours. During the study, acoustic Doppler velocimeters (ADV) were sampled at high frequency (50 Hz), while an acoustic Doppler current profiler (ADCP) and global positioning system (GPS) tracked drifters were used to obtain some lower frequency spatial distribution of the flow parameters within the estuary. The velocity measurements were complemented with some continuous measurement of water depth, conductivity, temperature and some other physiochemical parameters. Thorough quality control was carried out by implementation of relevant error removal filters on the individual data set to intercept spurious data. A triple decomposition (TD) technique was introduced to access the contributions of tides, resonance and ‘true’ turbulence in the flow field. The time series of mean flow measurements for both the ADCP and drifter were consistent with those of the mean ADV data when sampled within a similar spatial domain. The tidal scale fluctuation of velocity and water level were used to examine the response of the estuary to tidal inertial current. The channel exhibited a mixed type wave with a typical phase-lag between 0.035π– 0.116π. A striking feature of the ADV velocity data was the slow fluctuations, which exhibited large amplitudes of up to 50% of the tidal amplitude, particularly in slack waters. Such slow fluctuations were simultaneously observed in a number of physiochemical properties of the channel. The ensuing turbulence field showed some degree of anisotropy. For all ADV units, the horizontal turbulence ratio ranged between 0.4 and 0.9, and decreased towards the bed, while the vertical turbulence ratio was on average unity at z = 0.32 m and approximately 0.5 for the upper ADV (z = 0.55 m). The result of the statistical analysis suggested that the ebb phase turbulence field was dominated by eddies that evolved from ejection type process, while that of the flood phase contained mixed eddies with significant amount related to sweep type process. Over 65% of the skewness values fell within the range expected of a finite Gaussian distribution and the bulk of the excess kurtosis values (over 70%) fell within the range of -0.5 and +2. The TD technique described herein allowed the characterisation of a broader temporal scale of fluctuations of the high frequency data sampled within the durations of a few tidal cycles. The study provides characterisation of the ranges of fluctuation required for an accurate modelling of shallow water dispersion and mixing in a sub-tropical estuary.
Resumo:
A computer program has been developed for the prediction of buoyancy-driven laminar and turbulent flow in rectangular air-filled two-dimensional cavities with differentially heated side walls. Laminar flow predictions for a square cavity and Rayleigh numbers from Ra = 10^3 up to the onset of unsteady flow have been obtained. Accurate solutions for Ra = 5 x 10^6, 10^7, 5 x 10^7 and 10^8 are presented and an estimate for the critical Rayleigh number at which the steady laminar flow becomes unsteady is given for this geometry. Numerical predictions of turbulent flow have been obtained for RaH~0(10^9 -10^11 ) and compared with existing experimental data. A previously developed second moment closure model (Behnia et al. 1987) has been used to model the turbulence. Results indicate that a second moment closure model is capable of predicting the observed flow features.
Resumo:
"This chapter discusses laminar and turbulent natural convection in rectangular cavities. Natural convection in rectangular two-dimensional cavities has become a standard problem in numerical heat transfer because of its relevance in understanding a number of problems in engineering. Current research identified a number of difficulties with regard to the numerical methods and the turbulence modeling for this class of flows. Obtaining numerical predictions at high Rayleigh numbers proved computationally expensive such that results beyond Ra ∼ 1014 are rarely reported. The chapter discusses a study in which it was found that turbulent computations in square cavities can't be extended beyond Ra ∼ O (1012) despite having developed a code that proved very efficient for the high Ra laminar regime. As the Rayleigh number increased, thin boundary layers began to form next to the vertical walls, and the central region became progressively more stagnant and highly stratified. Results obtained for the high Ra laminar regime were in good agreement with existing studies. Turbulence computations, although of a preliminary nature, indicated that a second moment closure model was capable of predicting the experimentally observed flow features."--Publisher Summary
Resumo:
A computational model for isothermal axisymmetric turbulent flow in a quarl burner is set up using the CFD package FLUENT, and numerical solutions obtained from the model are compared with available experimental data. A standard k-e model and and two versions of the RNG k-e model are used to model the turbulence. One of the aims of the computational study is to investigate whether the RNG based k-e turbulence models are capable of yielding improved flow predictions compared with the standard k-e turbulence model. A difficulty is that the flow considered here features a confined vortex breakdown which can be highly sensitive to flow behaviour both upstream and downstream of the breakdown zone. Nevertheless, the relatively simple confining geometry allows us to undertake a systematic study so that both grid-independent and domain-independent results can be reported. The systematic study includes a detailed investigation of the effects of upstream and downstream conditions on the predictions, in addition to grid refinement and other tests to ensure that numerical error is not significant. Another important aim is to determine to what extent the turbulence model predictions can provide us with new insights into the physics of confined vortex breakdown flows. To this end, the computations are discussed in detail with reference to known vortex breakdown phenomena and existing theories. A major conclusion is that one of the RNG k-e models investigated here is able to correctly capture the complex forward flow region inside the recirculating breakdown zone. This apparently pathological result is in stark contrast to the findings of previous studies, most of which have concluded that either algebraic or differential Reynolds stress modelling is needed to correctly predict the observed flow features. Arguments are given as to why an isotropic eddy-viscosity turbulence model may well be able to capture the complex flow structure within the recirculating zone for this flow setup. With regard to the flow physics, a major finding is that the results obtained here are more consistent with the view that confined vortex breakdown is a type of axisymmetric boundary layer separation, rather than a manifestation of a subcritical flow state.
Resumo:
We report here a CFD model of highly swirling flow in a quarl burner using three versions of the k-epsilon model. Results for the recirculating zone, the bounding shear layer and the downstream flow are presented. We discuss, with suitable qualifications, how the model predictions can inform our understanding of this class of flows.