Research Papers: Flows in Complex Systems

J. Fluids Eng. 2019;141(9):091101-091101-15. doi:10.1115/1.4042885.

Through numerical simulation and experiments analysis, it is indicated that the hydraulic and anticavitation performance of a centrifugal pump with twisted gap drainage blades based on flow control theory can be significantly improved under certain operating conditions. In order to introduce the technology of gap drainage to practical applications, we put forward the parameter formulas of the twisted gap drainage blade to design three-dimensional new type blade, which are also proved to be effective for enhancing the dynamic characteristics of the centrifugal pump. Furthermore, a practical centrifugal pump is redesigned to be a twisted gap drainage impeller with the same structure size as the original impeller, and the nonlinear hybrid Reynolds-averaged Navier–Stokes (RANS)/large eddy simulation (LES) method is employed to simulate the hydraulic dynamic characteristics. Numerical simulation results show that the hydraulic performance and dynamic characteristics of the redesigned impeller centrifugal pump are significantly enhanced. In experiments, the twisted gap drainage blades structure not only remarkably improves the hydraulic performance and the pressure pulsation characteristics of the centrifugal pump but also reduces the vibration intensity.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091102-091102-16. doi:10.1115/1.4042755.

The bubble formation frequency from a single-orifice nozzle subjected to the effects of a crossflowing liquid was investigated using high-speed shadowgraphy, combined with image analysis and signal processing techniques. The effects of the nozzle dimensions, orientation within the conduit, liquid cross-flow velocity, and gas mass flow rate were evaluated. Water and air were the working fluids. Existing expressions in the literature were compared to the experimental values obtained. The expressions showed modest agreement with the experimental mean average frequency magnitude. It was found that increasing the gas injection diameter could decrease the bubbling frequency approximately 12% until reaching a certain value (0.52 mm). Further increasing the nozzle dimensions increase the frequency by around 20%. Bubbling frequency is more sensitive to the liquid velocity where changes up to 63% occurred when the velocity was raised from 3.1 to 4.3 m/s. Increasing gas mass flow rates decreased the gas jet breakup frequency in all cases. This phenomenon was primarily attributed to changes in the bubbling mode from discrete bubbling to pulsating and jetting modes. The nozzle orientation plays a role in modifying the bubbling frequency, having a higher magnitude when oriented against gravity.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091103-091103-11. doi:10.1115/1.4042884.

An experimental investigation has been conducted on rotating cavitation in a turbopump inducer. Previous research suggested that incidence angle variation leads to rotating cavitation. This study, using particle image velocimetry (PIV) method, provides the first measurement of the incidence angle distribution near the tip region of an inducer blade's leading edge with and without rotating cavitation. Without rotating cavitation, the incidence angle near the leading edge of the following blade is positive. Under rotating cavitation, the tip leakage vortex cavitation regions on the blades become uneven, forming large and small tip leakage vortex cavitation regions. A large tip leakage vortex cavitation on the leading blade increases the axial and absolute tangential velocities near the following blade's leading edge. Thus, the incidence angle on the following blade becomes negative, suppressing tip leakage vortex cavitation. Conversely, a small tip leakage vortex cavitation on the leading blade increases the incidence angle and promotes tip leakage vortex cavitation on the following blade. Due to such suppression and promotion mechanisms, the tip leakage vortex cavitation region on each blade oscillates in sequence, seemingly but not actually propagating in the forward direction. High-speed camera flow visualization has been used to confirm the same oscillation mechanism of rotating cavitation.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091104-091104-12. doi:10.1115/1.4042761.

In this paper, a deterministic stress decomposition is applied over the numerical three-dimensional flow solution available for a single volute centrifugal pump. The numerical model has proven in previous publications its robustness to obtain the impeller to volute-tongue flow interaction, and it is now used as starting point for the current research. The main objective has been oriented toward a detailed analysis of the lack of uniformity in the flow that the volute tongue promotes on the blade-to-blade axisymmetric pattern. Through this analysis, the fluctuation field may be retrieved and main interaction sources have been pinpointed. The results obtained with the deterministic analysis become of paramount interest to understand the different flow features found in a typical centrifugal pump as a function of the flow rate. Moreover, this postprocessing tool provides an economic and easy procedure for designers to compare the different deterministic terms, also giving relevant information on the unresolved turbulence intensity scales. Complementarily, a way to model the turbulent effects in a systematic way is also presented, comparing their impact on the performance with respect to deterministic sources in a useful framework, that may be applied for similar kinds of pumps.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091105-091105-28. doi:10.1115/1.4042752.

In this paper, capabilities of state-of-the-art computational fluid dynamics (CFD) tools in the prediction of the flow-field around a multihull catamaran advancing in straight ahead motion at nonzero drift angles are investigated. CFD estimations have been provided by three research institutes by using their in-house codes: CNR-INM using Xnavis, IIHR using CFDShip-Iowa, and CNRS/ECN using ISIS. These allowed an in-depth comparison between different methodologies, such as structured overlapping grids versus unstructured grid, different turbulence models and detached eddy simulations (DES) approaches, and level-set (LS) versus volume of fluid (VoF). The activities were pursued within the NATO AVT-183 group “reliable prediction of separated flow onset and progression for air and sea vehicles,” aimed at the assessment of CFD predictions of large three-dimensional separated flows. Comparison between estimations is provided for both integral and local quantities, and for wave-induced vortices. Validation is reported by comparison against the available experimental fluid dynamics (EFD) data. Generally, all the simulations are able to capture the main features of the flow field; grid resolution effects are dominant in the onset phase of coherent structures and turbulence model affects the dynamic of the vortices. Hydrodynamic loads are in agreement between the submissions with standard deviation of about 3.5% for the resistance prediction and about 7% for lateral force and yaw moment estimation. Wave-induced vortices are correctly captured by both LS and VoF approaches, even if some differences have been highlighted, LS showing well-defined and long life vortices.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091106-091106-15. doi:10.1115/1.4042757.

Models and experiments are developed to investigate how a small amount of gas can cause large rectified motion of a piston in a vibrated liquid-filled housing when piston drag depends on piston position so that damping is nonlinear even for viscous flow. Two bellows serve as surrogates for the upper and lower gas regions maintained by Bjerknes forces. Without the bellows, piston motion is highly damped. With the bellows, the piston, the liquid, and the two bellows move together so that almost no liquid is forced through the gaps between the piston and the housing. This Couette mode has low damping and a strong resonance: the piston and the liquid vibrate against the spring formed by the two bellows (like the pneumatic spring formed by the gas regions). Near this resonance, the piston motion becomes large, and the nonlinear damping produces a large rectified force that pushes the piston downward against its spring suspension. A recently developed model based on quasi-steady Stokes flow is applied to this system. A drift model is developed from the full model and used to determine the equilibrium piston position as a function of vibration amplitude and frequency. Corresponding experiments are performed for two different systems. In the two-spring system, the piston is suspended against gravity between upper and lower springs. In the spring-stop system, the piston is pushed up against a stop by a lower spring. Model and experimental results agree closely for both systems and for different bellows properties.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091107-091107-6. doi:10.1115/1.4042887.

High cavitating or supercavitating flows in fuel injector systems are crucial since they improve the mixing and the fuel atomization into combustion chambers, decreasing both fuel consumption and pollutant emissions. However, there is a lack of information regarding the required time to obtain high cavitating flows at the nozzle outlet, from the start of the injection pulse. In this work, a new method to quantify the time to get supercavitating flows at the nozzle outlet is developed. In particular, the delay in the inception of a supercavitating flow through a micronozzle is numerically analyzed for different pressure drops in a well-studied benchmark for fuel injectors. The three-dimensional simulations show that a delay higher than 100 μs is necessary for moderate pressure drops. Nevertheless, the delay tends to decay by rising amplitudes of the pressure pulse, reaching a saturation value of around 65 μs.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091108-091108-10. doi:10.1115/1.4043230.

Pressure flow generally exists in water conservancy projects and pipelines. The flow boundary of the contraction section faces a potential risk of cavitation erosion under high velocity. However, there is a lack of effective methods to suppress cavitation in engineering practices with pressure flow, posing a challenge to the operational safety of discharge structures and pipeline devices. The purpose of this paper was to realize the application of air entrainment in a plug-type contraction section of pressure flow. It was found that a single air vent and a low air flow rate could achieve complete vena contracta aeration. The pressure profiles of the vena contracta were investigated, and the results showed that the pressure distribution allowed the entrained air to diffuse laterally and convectively. Finally, we proposed a fitting algorithm to predict the air concentration in the vena contracta. These conclusions are of great significance for improving the safety and cavitation resistance of the contraction section of pressure flow.

Commentary by Dr. Valentin Fuster

Research Papers: Multiphase Flows

J. Fluids Eng. 2019;141(9):091301-091301-15. doi:10.1115/1.4042890.

Accurate simulation of the complex flow following the detonation of an explosive material is a challenging problem. In these flows, the detonation products of the explosive must be treated as a real gas while the surrounding air is treated as an ideal gas. As the detonation process unfolds and the blast wave moves into the surrounding ambient air, the products of detonation expand outward and interact with the air creating a mixture region. In this region, both of the state equations for air and the products must be satisfied. One of the most accurate, yet computationally expensive, methods to handle this problem is an algorithm that iterates between the equations of state until both pressure and temperature reach an equilibrium inside of a computational cell. Since this mixture region moves and grows over time, this algorithm must be performed millions, or even billions, of times in a typical detonation simulation. As such, these calculations can account for a large percentage of the overall solution time. This work aims to use a kriging surrogate model to replace this process. The iterative method solves a nonlinear system of equations created from the gas mixture density, internal energy, and composition using a Broyden iterative solver to obtain an output pressure and temperature. Kriging is used to produce curve fits which interpolate selected pressures and temperatures from this solver from appropriate ranges of the mixture input quantities. Using a finite volume hydrocode, the performance of the model with respect to the iterative solver is demonstrated in the simulation of a pentaerythritol tetranitrate (PETN) charge detonation. The model's computational speed and accuracy are quantified as a function of the choice of sampling points in order to try optimize the combination as well as to show the benefits of this novel approach.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091302-091302-20. doi:10.1115/1.4043055.

When an annular bed of solid particles that surrounds a cylindrical high-energy explosive core gets radially dispersed after detonation, the expanding front of particles undergoes instabilities. One of the possible causes of the instabilities is an inhomogeneous initial distribution of particles. This study explores this possibility by introducing two-dimensional perturbations to the initial distribution of particles within the annular bed and quantifying the growth of these perturbations over time using two-dimensional simulations. The initial perturbations are in the form of superposition of up to three sinusoidal azimuthal modal variations in the initial particle volume fraction (PVF, ratio of particle to cell volume). These are observed to impact the particle distribution at later times through a channeling instability whose effects are: (i) to decrease the velocity in regions of larger particle volume (PV) and (ii) to facilitate circumferential particle migration into the slow moving high PV sectors. These departures from axisymmetry are quantified by introducing two metrics. The effect of varying the number of azimuthal modes contained in the initial PVF perturbation, along with their amplitudes, wavelengths, and relative phases is investigated. The proposed metrics do not vary substantially with the relative phases; however, there is a strong variation in the metrics due to changes in the wavenumber. Unimodal perturbations were found to amplify both metrics the most.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091303-091303-10. doi:10.1115/1.4043161.

Prediction of the operational envelope (OE-limited range of gas and liquid velocities) for liquid carry-over is essential for the optimized performance of gas-liquid cylindrical cyclone (GLCC©) compact separators. This study presents for the first time the operational envelop for three-phase gas-oil-water flow incorporating pressure and level control configurations. A series of experiments were conducted to evaluate the performance of a 3 in. diameter GLCC in terms of OE for liquid carry-over. Experiments were carried out at different watercuts ranging from 0% to 100% utilizing water and two different types of mineral oils namely: light oil and heavy oil with specific gravities of 0.859 and 0.937, respectively. The liquid level was controlled at 6 in. below the GLCC inlet for all the experimental flow conditions. The experimental results indicate that OE for liquid carry-over for three-phase flow is very sensitive to watercut. As the watercut reduces, the OE for liquid carry-over reduces monotonically. Also, the OE for heavy oil (indicated by higher viscosity) reduces as compared to light oil. The superficial gas velocity required to create an annular mist flow in the upper part of the GLCC increases with the increase of watercut and viscosity.

Commentary by Dr. Valentin Fuster

Research Papers: Techniques and Procedures

J. Fluids Eng. 2019;141(9):091401-091401-8. doi:10.1115/1.4042886.

To characterize the microflow over a larger range of Knudsen numbers, an improved kinetic equation considering the volume diffusion effect for nonideal gases was presented based on Klimontovich's kinetic equation and Enskog equation-based lattice Boltzmann Bhatnagar–Gross–Krook (LBGK) model. Then, with the modified effective viscosity and the second-order slip boundary condition, a series of numerical simulations of gas flows with different mean Knudsen numbers were carried out based on the proposed model. Compared with the solutions of Navier–Stokes equations, Navier–Stokes equations with different slip boundary conditions, bivelocity hydrodynetics, and experimental data, we found that the present model can be valid up to a Knudsen number of 30. It is also shown that the present model furnishes a better solution in the transitional flow regime (0.1 < Kn < 10). The results not only illustrate that the present model could offer a satisfactory solution to a wider range of mean Knudsen number, but also show the importance of the compressibility and surface-dominated effects in micro gas flows. The improved model provides a promising tool for handling the micro gas flows with complex geometries and boundaries.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091402-091402-10. doi:10.1115/1.4042956.

Successful verification and validation is crucial to build confidence in the application of coupled computational fluid dynamics–discrete element method (CFD–DEM). Model verification includes ensuring a mesh-independent solution, which poses a major difficulty in CFD–DEM due to the complicated relationship between solution and computational cell size. In this paper, we investigate the production of numerical error in the CFD–DEM coupling procedure with computational grid refinement. The porosity distribution output from simulations of fixed-particle beds is determined to be Gaussian, and the average and standard deviation of the representative distribution are reported against cell size. We find that the standard deviation of bed porosity increases exponentially as the cell size is reduced. The average drag calculated from each drag law is very sensitive to changes in the porosity standard deviation. When combined together, these effects result in an exponential change in expected drag force when the cell size is small relative to the particle diameter. The divided volume fraction method of porosity calculation is shown to be superior to the centered volume fraction (CVF) method. The sensitivity of five popular drag laws to changes in the porosity distribution is presented, and the Ergun and Beetstra drag laws are shown to be the least sensitive to changes in the cell size. A cell size greater than three average particle diameters is recommended to prevent errors in the simulation results. A grid refinement study (GRS) is used to quantify numerical error.

Commentary by Dr. Valentin Fuster
J. Fluids Eng. 2019;141(9):091403-091403-18. doi:10.1115/1.4042963.

It is a well-known fact and a much studied problematic that the performance of low-head hydraulic turbines is highly dependent on the runner–draft tube coupling. Around the optimal operating conditions, the efficiency of the turbine follows closely the performance of the draft tube that in turn depends on the velocity field exiting the runner. Hence, in order to predict correctly the performance of the draft tube using numerical simulations, the flow inside the runner must be simulated accurately. Using results from unique and detailed particle image velocimetry (PIV) and laser Doppler velocimetry (LDV) measurements inside the runner channel of a bulb turbine, this paper presents an extensive study of the predictive capability of a widely used simulation methodology based on unsteady Reynolds-averaged Navier–Stokes equations with a k-epsilon closure model. The main objective was to identify the main parameters influencing the numerical predictions of the velocity field at the draft tube entrance in order to increase the accuracy of the simulated performance of the turbine. This paper relies on a comparison of simulations results with already published LDV measurements in the draft tube cone, interblade LDV, and stereoscopic PIV measurements within the runner. This paper presents a detailed discussion of numerical–experimental data correlation inside the runner channel and at the drat tube entrance. It shows that, contrary to widely circulated ideas, the near-wall predictions at the draft tube entrance is surprisingly good while the simulation accuracy inside the runner channels deteriorates from the leading to the trailing edges.

Commentary by Dr. Valentin Fuster

Technical Brief

J. Fluids Eng. 2019;141(9):094501-094501-7. doi:10.1115/1.4042667.

The investigation focused on the conversions of flow structures with a change in angle of attack (AOA) for a flexible cantilever wing, which experienced a self-excited vibration. Stereoscopic particle imaging velocimetry (Stereo-PIV) was utilized to measure the velocity field in the wing-tip region as AOA varied from 0 deg to 12 deg. At the Reynolds number (Re) of 3 × 104, instability waves shedding from the wing were amplified as they propagated and developed into Karman Vortex Street in the far downstream region at low AOAs (AOA = 4 deg and 6 deg). As AOA increased to 8 deg with the wing model was still steady, the Karman Vortex Street no longer existed. The wing started to vibrate at AOA = 10 deg owing to the self-excited vibration, and the Karman Vortex Street appeared again. The inception location of the Karman Vortex Street moved further upstream than in the cases at AOA = 4 deg and 6 deg. A new vortex structure, secondary vortex-pairs, appears outside the main wing-tip vortex (WTV).

Commentary by Dr. Valentin Fuster

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In