Stabilization of projection-based reduced order models of the Navier–Stokes

A new method of stabilizing low-order, proper orthogonal decomposition based reduced-order models of the Navier?CStokes equations is proposed. Unlike traditional approaches, this method does not rely on empirical turbulence modeling or modification of the Navier?CStokes equations. It provides spatial basis functions different from the usual proper orthogonal decomposition basis function in that, in addition to optimally representing the solution, the new proposed basis functions also provide stable reduced-order models. The proposed approach is illustrated with two test cases: two-dimensional flow inside a square lid-driven cavity and a two-dimensional mixing layer.     

A Newton/upwind method and numerical study of shock wave/boundary layer interactions

The objective of the paper is twofold. First we describe an upwind/central differencing method for solving the steady Navier–Stokes equations. The symmetric line relaxation method is used to solve the resulting algebraic system to achieve high computational efficiency. The grid spacings used in the calculations are determined from the triple-deck theory, in terms of Mach and Reynolds numbers and other flow parameters. Thus the accuracy of the numerical solutions is improved by comparing them with experimental, analytical and other computational results. Secondly we proceed to study numerically the shock wave/boundary layer interactions in detail, with special attention given to the flow separation. The concept of free interaction is confirmed. Although the separated region varies with Mach and Reynolds numbers, we find that the transverse velocity component behind the incident shock, which has not been identified heretofore, is also an important parameter. A small change of this quantity is sufficient to eliminate the flow separation entirely.     

Numerical investigation of a perturbed swirling annular two-phase jet

A swirling annular gas–liquid two-phase jet flow system has been investigated by solving the compressible, time-dependent, non-dimensional Navier–Stokes equations using highly accurate numerical methods. The mathematical formulation for the flow system is based on an Eulerian approach with mixed-fluid treatment while an adjusted volume of fluid method is utilised to account for the gas compressibility. Surface tension effects are captured by a continuum surface force model. Swirling motion is applied at the inlet while a small helical perturbation is also applied to initiate the instability. Three-dimensional spatial direct numerical simulation has been performed with parallelisation of the code based on domain decomposition. The results show that the flow is characterised by a geometrical recirculation zone adjacent to the nozzle exit and by a central recirculation zone further downstream. Swirl enhances the flow instability and vorticity and promotes liquid dispersion in the cross-streamwise directions. A dynamic precessing vortex core is developed demonstrating that the growth of such a vortex in annular configurations can be initiated even at low swirl numbers, in agreement with experimental findings. Analysis of the averaged results revealed the existence of a geometrical recirculation zone and a swirl induced central recirculation zone in the flow field.     

Explicit algebraic Reynolds stress model for shock-dominated flows

Shock waves drastically alter the nature of Reynolds stresses in a turbulent flow, and conventional turbulence models cannot reproduce this effect. In the present study, we employ explicit algebraic Reynolds stress model (EARSM) to predict the Reynolds stress anisotropy generated by a shockwave. The model by Wallin and Johansson (2000) is used as the baseline model. It is found to over-predict the post-shock Reynolds stresses in canonical shock turbulence interaction. The budget of the transport equation of Reynolds stresses computed using linear interaction analysis shows that the unsteady shock distortion mechanism and the pressure–velocity correlations are important. We propose improvement to the baseline model using linear interaction analysis results and redistribute the turbulent kinetic energy between the principle Reynolds stresses. The new model matches DNS data for the amplification of Reynolds stresses across the shock and their post-shock evolution, for a range of Mach numbers. It is applied to oblique shock/boundary-layer interaction at Mach 5. Significant improvements are observed in predicting surface pressure and skin friction coefficient, with respect to experimental measurements.     

Cover Image

A simple, robust, and accurate HLLC-type Riemann solver for the compressible Euler equations at various Mach numbers is built. To cure shock instability of the HLLC solver at strong shocks, a pressure-control technique, which plays a role in limiting the propagation of erroneous pressure perturbation, is proposed. With an all Mach correction method for the compressible Euler system, the proposed method is further extended to compute flow problems at low Mach numbers. The proposed all Mach HLLC-type scheme has been implemented and used to compute a variety of flow problems ranging from hypersonic compressible to low Mach incompressible flow regimes. Various numerical results demonstrate that the obtained all Mach HLLC-type scheme is both accurate and stable for all speed ranges.     

Shear-layer instability in the Mach reflection of shock waves

The Kelvin–Helmholtz instability (KHI) is an instability that takes the form of repeating wave-like structures which forms on a shear layer where two adjacent fluids are moving at a relative velocity to one another. Such a shear layer forms in the Mach reflection of shock waves. This work focuses on experimentally visualising the presence of the KHI in Mach reflection as well as its evolution. Experimentation was performed at shock Mach numbers of 1.34, 1.46 and 1.61. Plane test pieces and parabolic profiled pieces followed by a plane section having wedge angles of 30 \(^\circ \) and 38 \(^\circ \) were tested. Flow field visualisation was performed with a schlieren optical system. The KHI was best visualised with the camera-side knife edge perpendicular to the shear layer (i.e. the axis of sensitivity along the length of the shear layer). The structure and growth of the instability were readily identified. The KHI forms more readily with increasing Mach number and wedge angle. Second-order Euler, and Navier–Stokes numerical simulations of the flow field were also conducted. It was found that the Euler and laminar Navier–Stokes solvers achieved very similar results, both producing the KHI, but at a much less developed state than the experimental cases. The k \(-\epsilon \) solver, however, did not produce the instability.     

A criterion for detection of the onset of Dean instability in Newtonian fluids

Dean instability for Newtonian fluids in laminar secondary flow in 180° curved channels was studied experimentally and numerically. The numerical study used Fluent CFD code to solve the Navier–Stokes equations, focusing on flow development conditions and the parameters influencing Dean instability. An accurate criterion based on the radial gradient of the axial velocity was defined that allows detection of the instability threshold, and this criterion is used to optimize the grid geometry. The effects on Dean instability of the curvature ratio (from 5.5 to 20) and aspect ratio (from 0.5 to 12) are studied. In particular, we show that the critical value of the Dean number decreases with the increasing duct curvature ratio. The variation of the critical Dean number with duct aspect ratio is less regular.In the experimental study, flows were visualized in several tangential positions of a 180° curved channel with aspect ratio 8 and curvature ratio 10. The flow is hydrodynamically developed at the entrance to the curved channel. The critical Dean number is detected and the development of secondary flow vortices by additional counter-rotating vortex pairs is observed. A diagram of different critical Dean numbers is established.     

Response of plane viscous jet to entrance flow rate perturbation

A two‐dimensional numerical simulation solving unsteady incompressible Navier–Stokes equations is used to study the natural varicose instability of a plane jet in the Reynolds number range of 100–900. A transient train of vortices is observed at the beginning of the computation. It disappears yielding a steady flow. This flow is then used at the basis for forced excitation in order to study the space time development of instability. A Reynolds number dependant behaviour is observed which implies that viscosity directly affects the vortex dynamics. Copyright © 2001 John Wiley & Sons, Ltd.     

The role of vortex wake dynamics in the flow-induced vibration of tube arrays

Potential flow and 2-D Navier–Stokes calculations are used to investigate the role of vortex shedding in the non-resonant flow-induced vibration of periodic tube arrays. This dual approach untangles the effects of potential and vortical flow. The negative damping theory is shown to be inconsistent with the Navier–Stokes simulations, and allowing only a single degree of freedom in tube motion significantly overestimates the critical velocity. In contrast, Navier–Stokes simulations which allow all tubes to move in both the transverse and streamwise directions give results in good agreement with experiment. Somewhat surprisingly, potential flow calculations including an artificial phase lag between fluid force and tube motion give reasonably accurate results for a wide range of phase lags. This may be due to the fact that the most unstable mode at onset appears to be streamwise anti-phase (not whirling), as observed in the potential flow case.     

Viscous solution of the triple-shock reflection problem

The reflection of a triple-shock configuration was studied numerically in two dimensions using the Navier–Stokes equations. The flow field was initialized using three shock theory, and the reflection of the triple point on a plane of symmetry was studied. The conditions simulated a stoichiometric methane-oxygen detonation cell at low pressure on time scales preceding ignition when the gas was assumed to be inert. Viscosity was found to play an important role on some shock reflection mechanisms believed to accelerate reaction rates in detonations when time scales are small. A small wall jet was present in the double Mach reflection and increased in size with Reynolds number, eventually forming a small vortex. Kelvin–Helmholtz instabilities were absent, and there was no Mach stem bifurcation at Reynolds numbers corresponding to when the Mach stem had travelled distances on the scale of the induction length. Kelvin–Helmholtz instabilities are found to not likely be a source of rapid reactions in detonations at time scales commensurate with the ignition delay behind the Mach stem.     

A 3D UNSTRUCTURED MESH ADAPTATION ALGORITHM FOR TIME-DEPENDENT SHOCK-DOMINATED PROBLEMS

In this paper we present a tetrahedron-based, h-refinement-type algorithm for the solution of problems in 3D gas dynamics using unstructured mesh adaptation. The mesh adaptation algorithm is coupled to a cell-centred, Riemann problem-based, finite volume scheme of the MUSCL type, employing an approximate Riemann solver. The adaptive scheme is then used to compute the diffraction of shock waves around a box section corner for subsonic and supersonic post-shock flow. In the subsonic case, preliminary measurements of vortex filament speed and vortical Mach number are in broad quantitative agreement with known theoretical results. © 1997 John Wiley & Sons, Ltd.     

Stochastic Three-Dimensional Rotating Navier–Stokes Equations: Averaging, Convergence and Regularity

We consider stochastic three-dimensional rotating Navier?CStokes equations and prove averaging theorems for stochastic problems in the case of strong rotation. Regularity results are established by bootstrapping from global regularity of the limit stochastic equations and convergence theorems.     

Simplified system of equations for calculating the stability of the boundary layer gas flow on a swept wing

A composite system of equations for calculating the stability of a compressible flow in a three-dimensional boundary layer is developed. A comparison with the calculations in accordance with the complete system of equations shows that the system proposed almost accurately predicts the growth rates of both the Tollmien-Schlichting mode and the transverse flow instability mode over a wide range of frequencies and transverse wavenumbers at moderate Mach numbers. At the same time, the computation time reduces by an order.     

Effect of incident shock wave strength on the decay of Richtmyer–Meshkov instability-introduced perturbations in the refracted shock wave

The effect of incident shock wave strength on the decay of interface introduced perturbations in the refracted shock wave was studied by performing 20 different simulations with varying incident shock wave Mach numbers (M ~ 1.1? 3.5). The analysis showed that the amplitude decay can be represented as a power law model shown in Eq.7, where A is the average amplitude of perturbations (cm), B is the base constant (cm^?(E?1), S is the distance travelled by the refracted shockwave (cm), and E is the power constant. The proposed model fits the data well for low incident Mach numbers, while at higher mach numbers the presence of large and irregular late time oscillations of the perturbation amplitude makes it hard for the power law to fit as effectively. When the coefficients from the power law decay model are plotted versus Mach number, a distinct transition region can be seen. This region is likely to result from the transition of the post-shock heavy gas velocity from subsonic to supersonic range in the lab frame. This region separates the data into a high and low Mach number region. Correlations for the power law coefficients to the incident shock Mach number are reported for the high and low Mach number regions. It is shown that perturbations in the refracted shock wave persist even at late times for high incident Mach numbers.     

Experimental investigation of vortex formation in the wake of a flat plate for subsonic and supersonic freestream Mach numbers

The wake of a flat plate with thickness H = 6 mm whose surface was either smooth or rough by pasting smooth or rough paper on to it was investigated with optical methods. The studies showed that for supersonic freestream Mach numbers the wake of the smooth flat plate seems to be turbulent in the observation field. In contrast the wake of the rough plate can exhibit a vortex street if certain conditions are met. The most important parameters are the roughness of the paper and the distance from the location where the rough paper ends to the trailing edge, abbreviated with d. It turned out that a vortex street develops in the wake if the roughness exceeds a critical value and if the distance d is of the order of some millimeters. If these conditions are not met the wake of the rough plate is turbulent in the observation field. The dependence of the vortex formation upon the Mach number, the roughness and the distance d was investigated with holographic interferometry, two component laser-Doppler-anemometry and a laser-optics for measuring the vortex shedding frequency.     

DNS of flows with helical symmetry

Some flows such as the wakes of rotating devices often display helical symmetry. We present an original DNS code for the dynamics of such helically symmetric systems. We show that, by enforcing helical symmetry, the three-dimensional Navier–Stokes equations can be reduced to a two-dimensional unsteady problem. The numerical method is a generalisation of the vorticity/streamfunction formulation in a circular domain, with finite differences in the radial direction and spectral decomposition along the azimuth. When compared to a standard three-dimensional code, this allows us to reach larger Reynolds numbers and to compute quasi-steady patterns. We illustrate the importance of helical pitch by some physical cases: the dynamics of several helical vortices and a quasi-steady vortex flow. We also study the linear dynamics and nonlinear saturation in the Batchelor vortex basic flow, a paradigmatic example of trailing vortex instability. We retrieve the behaviour of inviscid modes and present new results concerning the saturation of viscous centre modes.     

Self-intensification in shock wave and vortex interaction

Computer simulation has been performed for the interaction between a shock wave and a vortex ring moving toward the wave. The computed density contours are compared with the pattern of shadowgraphs. A remarkable property found in the simulation is that, during the passage of the shock wave over the vortex ring, the part of the wave propagating through the inside of the ring-vortex is intensified spontaneously at a localized region. Maximum pressure occurs inside the vortex and attains a high value, about several times that of the impinging shock for incident Mach numbers of around 1.2 with the vortex translation Mach number 0.60. This is due to a double-step mechanism of intensification within the flow field by the shock-vortex interaction.     

Nonlinear Strong Shock Interactions: A Shock-Fitted Approach

This paper addresses nonlinear effects which result from the interaction of shock waves with vortices. A series of experiments are carried out, which involve the interaction of a strong shock wave with a single plane vorticity wave and a randomly distributed wave system. These experiments are first conducted in the linear regime to obtain a mutual verification of theory and computation. They are subsequently extended into the nonlinear regime. A systematic study of the interaction of a plane shock wave and a single vortex is then conducted. Specifically, we investigate the conditions under which nonlinear effects become important, both as a function of shock Mach number, M ₁, and incident vortex strength (characterized by its circulation Γ). The shock Mach number is varied from 2 to 8, while the circulation of the vortex is varied from infinitesimally small values (linear theory) to unity. Budgets of vorticity, dilatation, and pressure are obtained. They indicate that nonlinear effects become more significant as both the shock Mach number and the circulation increase. For Mach numbers equal to 5 and above, the dilatation in the vortex core grows quadratically with circulation. An acoustic wave propagates radially outward from the vortex center. As circulation increases, its upstream-facing front steepens at low Mach numbers, and its downstream-facing front steepens at high Mach numbers. A high Mach number asymptotic expansion of the Rankine--Hugoniot conditions reveals that nonlinear effects dominate both the shock motion and the downstream flow for ΓM ₁ > 1. Received 28 June 1997 and accepted 25 November 1997     

Large Deborah number flows around confined microfluidic cylinders

Viscoelastic flow around a confined cylinder at high Deborah numbers is studied using microfluidic channels. By varying fluid properties and flow rates, a systematic study of the roles of elasticity and inertia is accomplished. Two new elastic flow instabilities that occur at high Deborah numbers are identified. A downstream instability of disordered and temporally varying streamlines is observed at a Deborah number above 10. This instability is a precursor to an unsteady vortex that develops upstream of the cylinder at higher Deborah numbers. Both instabilities occur at moderate Reynolds numbers but are fundamentally elastic. The size and steadiness of the upstream vortex are primarily controlled by the Deborah and the elasticity number.     

Leading-Order Dynamics of Gravity-Driven Flows in?Free-Standing Soap Films

We derive the leading-order equations that govern the dynamics of the flow in a falling, free-standing soap film. Starting with the incompressible Navier?CStokes equations, we carry out an asymptotic analysis using parameters that correspond to a common experimental setup. We account for the effects of inertia, surface elasticity, pressure, viscous stresses, gravity, and air drag. We find that the dynamics of the flow is dominated by the effects of inertia, surface elasticity, gravity, and air drag. We solve the leading-order equations to compute the steady-state profiles of velocity, thickness, and pressure in an experiment in which the film is in the Marangoni elasticity regime. The computational results, which include a Marangoni shock, are in good accord with the experimental measurements.