Firat Oguz Edis

Computing Micro Synthetic Jet in Slip Regime With Moving Membrane

A Navier-Stokes (NS) solver for moving and deforming meshes has been modified to investigate numerically the diaphragm-driven flow in and out of two synthetic jet cavity geometries. The piezoelectric-driven diaphragm of the cavity is modeled in a realistic manner as a moving boundary to accurately compute the flow inside the jet cavity. The primary focus of the present paper is to describe the effect of cavity geometry and the wall slip, resulting from the relatively larger Kn number flows associated with micro sized geometries, on the exit jet velocity magnitude. Compressible flow simulations are required for rarefied flows to accurately predict the pressure field. The present computations for the quiescent external flow condition reveal that cavity geometry and the wall slip has an increasing effect on the magnitude of the average jet exit velocity as well as vortex shedding from the orifice.Copyright

Computational Investigation of Micro Bacward-Facing Step Duct Flow in Slip Regime

A comprehensive computational analysis is performed to investigate the nature of rarefied gas flow through a backward-facing step duct. A characteristic-based split Navier-Stokes FEM solver is used for the simulations of the flows found in slip regime. The second-order slip-velocity and temperature-jump boundary conditions of Beskok and Karniadakis are applied on the duct walls. Changing flow and heat transfer behavior of the flow due to the varying conditions characterized with non-dimensional parameters are presented and discussed in this article.

Efficient incompressible flow calculations using PQ2Q1 element

Implementation of an equal-order-interpolation velocity–pressure element pair is presented for the finite element solution of incompressible viscous flows. A fractional-step method is employed for temporal discretization. The element pair, also called a pseudo-biquadratic velocity/bilinear pressure element (pQ2Q1), consists of a bilinear pressure element and bilinear velocity elements defined on subdivisions of the pressure element. This pair satisfies the so-called ‘Ladyzhenskaya–Babuska–Brezzi’ condition. Considerable savings in computational cost are achieved due to the reduced number of elements for pressure. A modification of the element is realized for a better representation of curved surfaces. Two test cases, namely the lid-driven cavity flow and impulsively started circular cylinder in cross-flow, are used to assess the accuracy and efficiency of the element compared to a regular bilinear velocity–pressure (Q1Q1) element pair. Computational results presented show that the pQ2Q1 element solutions require less memory and CPU time compared to Q1Q1 element solutions, for at least the same accuracy.

Micro‐scale synthetic‐jet actuator flow simulation with characteristic‐based‐split method

Purpose – The purpose of this paper is to present a computational study to investigate the effects of rectangular cavity design of a piezoelectrically driven micro‐synthetic‐jet actuator on generated flow.Design/methodology/approach – Flow simulations were done using a compressible Navier‐Stokes solver, which is based on finite element method implementation of a characteristic‐based‐split (CBS) algorithm. The algorithm uses arbitrary Lagrangian‐Eulerian formulation, which allows to model oscillation of the synthetic jets diaphragm in a realistic manner. Since all simulated flows are in the slip‐flow‐regime, a second order slip‐velocity boundary condition was applied along the cavity and orifice walls. Flow simulations were done for micro‐synthetic‐jet configurations with various diaphragm deflections amplitudes, cavity heights, and widths. All of the simulation results were compared with each other and evaluated in terms of the exit jet velocities, slip‐velocities on the orifice wall and instantaneous mo...

Numerical simulation of Kelvin-Helmholtz instability using an implicit, non-dissipative DNS algorithm

An in-house, fully parallel compressible Navier-Stokes solver was developed based on an implicit, non-dissipative, energy conserving, finite-volume algorithm. PETSc software was utilized for this purpose. To be able to handle occasional instances of slow convergence due to possible oscillating pressure corrections on successive iterations in time, a fixing procedure was adopted. To demonstrate the algorithms ability to evolve a linear perturbation into nonlinear hydrodynamic turbulence, temporal Kelvin-Helmholtz Instability problem is studied. KHI occurs when a perturbation is introduced into a system with a velocity shear. The theory can be used to predict the onset of instability and transition to turbulence in fluids moving at various speeds. In this study, growth rate of the instability was compared to predictions from linear theory using a single mode perturbation in the linear regime. Effect of various factors on growth rate was also discussed. Compressible KHI is most unstable in subsonic/transonic regime. High Reynolds number (low viscosity) allows perturbations to develop easily, in consistent with the nature of KHI. Higher wave numbers (shorter wavelengths) also grow faster. These results match with the findings of stability analysis, as well as other results presented in the literature.

Short Note: Highly efficient volume generation reservoirs in molecular simulations of gas flows

DSMC (Direct Simulation Monte Carlo) method pioneered by Bird is a stochastic molecular simulation method that can be used for simulations of rarefied gas flows [1]. In DSMC method, each DSMC molecule represents a large number of physical gas molecules. DSMC molecules carry position, velocity and, if applicable, internal energy information on them. Molecule movements and collisions are decoupled from each other. Moving molecules either fly freely or interact with boundaries. In the collision step, translational and internal energies are re-shared between molecules according to the collision model chosen. Molecules interact either with wall or with stream boundaries. In case of wall boundaries, molecules are reflected back according to the reflection model chosen. When molecules cross stream boundaries, they leave the domain without any further interaction. At the same time new molecules are introduced into the flow area. The number of molecules introduced into the gas flow area and their velocity components depend on the boundary conditions. In the case of Pressure Boundary Conditions (PBC), temperature ðTiÞ, pressure ðpiÞ and stream velocity components parallel to the boundary ðVi;WiÞ are predetermined properties on upstream boundaries. On the downstream boundary only the pressure ðpoÞ is predetermined. All missing boundary conditions can be calculated with extrapolation from the gas flow area [2]. Particle flux conservation technique by [3] can also be used to calculate the stream velocity components normal ðUi;UoÞ to the stream boundaries. Molecule number density ðnÞ at the boundary can be calculated from the perfect gas equation of state. n 1⁄4 p kBT ð1Þ

Atomistic simulation of micro‐scale adiabatic piston problem

Purpose – The purpose of this paper is to demonstrate the utilization of the direct simulation Monte Carlo (DSMC) method for moving‐boundary/deforming‐domain micro‐scale gas flow problems. Furthermore, a hydrodynamic model, proposed in the literature, is used to compare its results with those obtained using the DSMC method.Design/methodology/approach – A micro‐scale adiabatic piston problem is analyzed using a parallel DSMC implementation for deforming domains. Initially, pressures at both sides of the piston wall are different. Consequently, frictionless piston moves toward low‐pressure compartment, keeps oscillating from one side to the other. Eventually, the piston reaches the “Mechanical equilibrium” state. Although the temperatures are different, pressures are equal at this state. The unsteady problem is analyzed until it reaches this state. Three test cases, all with the same initial conditions but different piston masses are analyzed. The time variation of the piston position, conditions in the com...

An implementation of the direct-forcing immersed boundary method using GPU power

ABSTRACT A graphics processing unit (GPU) is utilized to apply the direct-forcing immersed boundary method. The code running on the GPU is generated with the help of the Compute Unified Device Architecture (CUDA). The first and second spatial derivatives of the incompressible Navier-Stokes equations are discretized by the sixth-order central compact finite-difference schemes. Two flow fields are simulated. The first test case is the simulated flow around a square cylinder, with the results providing good estimations of the wake region mechanics and vortex shedding. The second test case is the simulated flow around a circular cylinder. This case was selected to better understand the effects of sharp corners on the force coefficients. It was observed that the estimation of the force coefficients did not result in any troubles in the case of a circular cylinder. Additionally, the performance values obtained for the calculation time for the solution of the Poisson equation are compared with the values for other CPUs and GPUs from the literature. Consequently, approximately 3× and 20× speedups are achieved in comparison with GPU (using CUSP library) and CPU, respectively. CUSP is an open-source library for sparse linear algebra and graph computations on CUDA.

Application of an All-Speed Implicit Finite-Volume Algorithm to Rayleigh–Taylor Instability

We present a three-dimensional Direct Numerical Simulation (DNS) study of Rayleigh–Taylor Instability (RTI) using an all-speed, fully implicit, nondissipative and discrete kinetic energy conserving algorithm. In order to perform this study, an in-house, fully parallel, finite-volume, DNS solver, iDNS, which solves the set of time-dependent, compressible Navier–Stokes equations with gravity was developed based on the present algorithm and the PETSc parallel library. It is shown that the algorithm is able to capture the correct physics of the baroclinic instability and turbulent mixing. Compressibility (i.e., high Mach number) has been found more effective on the development of the flow after the diffusive growth phase passed. An increase in bubble growth rate together with a decrease in turbulent mixing was also observed at Mach number 1.1.

Parallel implicit DNS of temporally-evolving turbulent shear layer instability

In this study, a temporally-evolving incompressible and compressible Turbulent Shear Layer (TSL) instability problem is solved using an all-speed (all-Mach), implicit, non-dissipative and kinetic energy conserving algorithm. An in-house, fully parallel, finite-volume Direct Numerical Simulation (DNS) solver was developed using PETSc. Convergence characteristics at low-Mach numbers were also improved using a relaxation procedure. We aim here to assess the performance and behavior of the present algorithm for complex flows which contain multi-scale physics and gradually evolve into turbulence. The results show that the algorithm is able to produce correct physical mechanisms and capture the evolution of the turbulent fluctuations for both incompressible and compressible cases. It is observed that the non-dissipative and kinetic energy conserving properties make the algorithm powerful and applicable to challenging problems. For higher Mach numbers, a shock-capturing or a dissipative mechanism is required for robustness.