Search results

A numerical study is performed to investigate flow instability phenomena in a square channel with steady, laminar throughflow. The channel rotates around an axis perpendicular to the channel longitudinal axis. The flow field extends from the channel entrance to a distance of 120 to 600Dh. The range of Reynolds number is Re = 300−2000. The inlet flow velocity is assumed uniform. Surface vorticity intensity is introduced to indicate the variation of vortices. It is revealed that at intermediate Reynolds numbers (680 > Re > 300), the flow is characterized by three vortex patterns: at slow rotation there is one vortex pair; at intermediate rotation a secondary vortex, in addition to the original vortex, emerges near the trailing wall and then breaks down downstream; and at rapid rotation the secondary vortex does not exist with the flow being restabilized to form a single‐pair vortex pattern. At low Reynolds numbers (Re ≤ 300), the flow exhibits a single‐pair vortex pattern, while at high Reynolds numbers (Re ≥ 680), the flow experiences the emergence and breakdown of a secondary vortex, but no restabilization is found with an increase in the rotational speed. It is also disclosed that the variation of the vortices is related to the distance from the inlet.

The purpose of this paper is to study the flow patterns and particle-particle collisions during the sedimentation of multiple circular particles under gravity at intermediate Reynolds numbers through direct numerical simulations (DNS).

The previously developed lattice Boltzmann-direct forcing/fictitious domain (LB-DF/FD) method is adopted in this work to conduct DNS.

It is found that the number of particle-particle collisions display a linear growth at long times after an initial evolution, resulting in a constant collision rate, which also depends the initial arrangement.

The problem of particle-particle collisions during sedimentation with two kinds of particle density has not been considered before and it is of special importance in various industries.

A numerical investigation aimed at understanding the flow and heat transfer characteristics of pulsating turbulent flow in an abrupt pipe expansion was carried out. The flow patterns are classified by four parameters; the Reynolds number, the Prandtl number, the abrupt expansion ratio and the pulsation frequency. The influence of these parameters on the flow was studied in the range 10⁴<Re<5×10⁴, 0.7<Pr<7.0, 0.2<d/D<0.6 and 5<f<35. It was found that the influence of pulsation on the mean time‐averaged Nusselt number is insignificant (around 10 per cent increase) for fluids having a Prandtl number less than unity. This effect is appreciable (around 30 per cent increase) for fluids having Prandtl number greater than unity. For all pulsation frequencies, the variation in the mean time‐averaged Nusselt number, maximum Nusselt number and its location with Reynolds number and diameter ratio exhibit similar characteristics to steady flows.

Upward mixed convection flow of air in a uniformly heated vertical tube was studied numerically using the three‐dimensional elliptic conservation equations and the Launder and Sharma low Reynolds number k–ε turbulence model. For Re=1,000 the fully developed flow field undergoes two transitions as the Grashof number increases: thus, this flow field is laminar for Gr<8×10⁶, turbulent for 8×10⁶<Gr<5×10⁷ and again laminar for Gr>5×10⁷. In the entry region, turbulent kinetic energy decays monotonically for Gr≤3×10⁶ and Gr≥7.1×10⁷. For Gr between these two values it initially increases from the imposed inlet condition and then decreases towards its calculated fully developed value. The mean velocity profiles as well as the axial evolution of the skin friction coefficient are presented for representative values of Gr.

A numerical study of heat transfer enhancement due to the deformation of droplets at high Reynolds numbers is described. The two phase‐flow has been computed with a 3D DNS program using the volume‐of‐fluid method. The droplets are deformed because of the surrounding gas stream especially due to a sudden rise of flow velocity from zero to U_i. As the governing non‐dimensional parameter the Weber number of the droplets has been varied between 1.3 and 10.8 by assuming different surface tensions at Reynolds numbers between 360 and 853. The dynamical behavior of the droplets as a function of the Weber and the Ohnsorge number are in good agreement with experimental results from the literature. At the highest Reynolds number Re=853, a significant dependency of Nu on We has been found. The comparison of a Nusselt number computed with the real surface area with a Nusselt number computed with the spherical surface area shows that the heat transfer increases not only due to the droplet motion but also due to the larger surface area of the deformed droplet.

To numerically model forced convection heat transfer over arrays of solder balls.

The characteristic based split (CBS) scheme has been used to solve the incompressible Navier‐Stokes equations on unstructured meshes.

The results show an increase in heat transport with increase in Reynolds numbers. A significant change in heat transfer is also noticed with change in angle of attack.

The presented results will be useful in designing cooling systems for electronic components.

The main purpose of this paper was to investigate Lattice Boltzmann (LB) models for the bulk incompressible flow past immersed bodies and to find the set of boundary conditions (BCs) that can be considered suitable for modeling the borders of the numerical simulation domain in such a way as to avoid any effect of these BC on the flow trail that is formed behind the body.

Three different models of the Lattice Boltzmann equation (LBE) and six different sets of BCs are tested. In addition to the classical LBE based on the Bhatnagar–Gross–Krook (BGK) single relaxation time collision model, a moments-based model and a model with two relaxation times were investigated.

The flow pattern and its macroscopic effects on the aerodynamic coefficients appear to be very dependent on the set of BC models used for the borders of the numerical domain. The imposition of pressure at the exit results in pressure perturbations, giving rise to sound waves that propagate back into the simulation domain, producing perturbations on the upwind flow. In the same way, the free-slip BC for the lateral bords appears to affect the trail of vortices behind the body in this range of Reynolds number (Re = 1,000).

The paper investigates incompressible flow past immersed bodies and presents the set of BCs that can be considered suitable for modeling the borders of the numerical simulation domain in such a way as to avoid any effect of these BCs on the flow trail that is formed behind the body.

Although many studies have been conducted on the nanofluid flow in microtubes, this paper, for the first time, aims to investigate the effects of nanoparticle diameter and concentration on the velocity and temperature fields of turbulent non-Newtonian Carboxymethylcellulose (CMC)/copper oxide (CuO) nanofluid in a three-dimensional microtube. Modeling has been done using low- and high-Reynolds turbulent models. CMC/CuO was modeled using power law non-Newtonian model. The authors obtained interesting results, which can be helpful for engineers and researchers that work on cooling of electronic devices such as LED, VLSI circuits and MEMS, as well as similar devices.

Present numerical simulation was performed with finite volume method. For obtaining higher accuracy in the numerical solving procedure, second-order upwind discretization and SIMPLEC algorithm were used. For all Reynolds numbers and volume fractions, a maximum residual of 10⁻⁶ is considered for saving computer memory usage and the time for the numerical solving procedure.

In constant Reynolds number and by decreasing the diameter of nanoparticles, the convection heat transfer coefficient increases. In Reynolds numbers of 2,500, 4,500 and 6,000, using nanoparticles with the diameter of 25 nm compared with 50 nm causes 0.34 per cent enhancement of convection heat transfer coefficient and Nusselt number. Also, in Reynolds number of 2,500, by increasing the concentration of nanoparticles with the diameter of 25 nm from 0.5 to 1 per cent, the average Nusselt number increases by almost 0.1 per cent. Similarly, In Reynolds numbers of 4,500 and 6,000, the average Nusselt number increases by 1.8 per cent.

The numerical simulation was carried out for three nanoparticle diameters of 25, 50 and 100 nm with three Reynolds numbers of 2,500, 4,500 and 6,000. Constant heat flux is on the channel, and the inlet fluid becomes heated and exists from it.

The authors obtained interesting results, which can be helpful for engineers and researchers that work on cooling of electronic devices such as LED, VLSI circuits and MEMS, as well as similar devices.

This manuscript is an original work, has not been published and is not under consideration for publication elsewhere. About the competing interests, the authors declare that they have no competing interests.

WHERE computer programmes are not available for the solution of kinetic heating problems, the estimation of equilibrium and transient surface temperatures on bodies in supersonic flow becomes tedious, especially where the transient conditions are to be analysed. A current method of performing these calculations for a turbulent boundary layer is here synthesised into nomogram form.

The conception of a heat function, just like the stream function used in a laminar two dimensional incompressible flow field visualization, has been introduced to visualize the convective heat transfer or the flow of energy around a sphere when the sphere is either being cooled or heated by a stream of fluid flowing around it. The heat function is developed in a spherical polar coordinate and is used to generate the heat lines around the sphere. The heat lines essentially show the magnitude and direction of energy transfer around the sphere with and without the existence of a finite radial velocity at the surface. The steady state hydrodynamic field around the sphere is numerically obtained up to a maximum Reynolds number of 100 and the corresponding thermal field has been obtained by solving the steady state energy equation. The field properties thus obtained are utilized to form the heat function, which becomes an effective tool for visualization of convective heat transfer.