Large Eddy Simulation Turbulence Model Research Articles

CFD analysis of combustion and emission formation using URANS and LES under large two-stroke marine engine-like conditions

In this paper, the main aim is to examine the performance of turbulence models to shed light on the effect of turbulence modeling in capturing different in-cylinder phenomena under large two-stroke marine engine-like conditions. The Unsteady Reynolds Averaged Navier–Stokes (URANS) and Large Eddy Simulation (LES) turbulence models are utilized. The LES and URANS results are compared with experimental data, in which LES and URANS models show similar accuracy in capturing the pressure and heat release with a moderately better accuracy in the LES case. The predicted gas temperature at the liner wall is approximately 45% higher for URANS than LES during the expansion stroke, which may lead to different sulfuric acid formation and heat transfer prediction. The LES model predicts a 34% higher average swirl than that in the URANS case which leads to an earlier and a stronger interaction between the flame and the spray, decreasing the oxidation of the emissions. Due to the higher predicted in-cylinder temperature in the LES case, the NO emission amount at exhaust valve opening time (EVO) is 7% higher in the LES case. At EVO, the CO emission in the LES case is predicted to be 3-fold higher than that in the URANS case due to less oxidation of CO in the post oxidation stage in the LES case. The second cycle LES simulation shows that the solutions after the scavenging process are in-sensitive to the initial conditions.

Dynamic Characteristics and Effect on Thrust of Jet Tail Vortex Ring for Pump Jet Propulsion

When a pump is propelled by a propeller, the nozzle flooding of the jet wake area will produce a turbulent quasi-sequence structure and have a certain impact on the outflow field structure and thrust characteristics of the water jet propulsion pump. In this paper, a method that combined numerical simulation with vortex dynamics is adopted, which analyzes the dynamic characteristics and influence on the thrust characteristics of the water jet propulsion pump. A large eddy simulation turbulence model and a dimensionless water jet propulsion pump velocity coefficient were used to reveal flow structure and relation, with the pump operation parameters of the wake vortex ring. The thrust with a trailing vortex ring is 7.0% higher than that without a trailing vortex ring. Vortex dynamics and mathematical statistics are combined to quantitatively analyze the dynamic characteristics of the jet tail vortex ring. Finally, the formation time of the vortex ring is obtained in exponential relation with dimensionless transmission velocity and vorticity coefficient, which has nonlinear relation with vortex intensity coefficient and helicity coefficient. BP Neural Network combined with the LM algorithm is used to establish the mathematical relationship between the thrust and the physical characteristic parameters of the vortex ring.

Aerodynamic characteristics of two noise barriers (fully enclosed and semi-enclosed) caused by a passing train: A comparative study

With the wide use of noise barriers along high-speed railways, the aerodynamic problems of noise barrier caused by high-speed train operation become increasingly prominent. At present, the commonly used noise barrier types of high-speed railway can be classified as fully enclosed and semi-closed (inverted ‘L’ type). The model test of high-speed moving train with scale ratio of 1:16.8 was conducted, and the temporal and spatial laws of aerodynamic pressure of fully enclosed and semi-enclosed noise barriers were compared. These processes were performed to study the difference in the aerodynamic pressure characteristics of noise barriers when high-speed trains at speeds of 350 km/h pass through the two types of noise barriers. On the basis of large eddy simulation turbulence model and ‘mosaic’ grid technology, the corresponding 3D computational fluid dynamics numerical model of train–noise barrier–bridge was established, and the corresponding improvement schemes were proposed to solve the phenomenon of unreasonable pulsation pressure distribution of fully enclosed and semi-enclosed noise barriers. The pressure relief mechanism of the pressure relief hole on the fully enclosed noise barrier and the buffer mechanism of the buffer structure at the end of the semi-enclosed noise barrier are revealed from the perspective of flow field. The main results show that: (1) Compression wave and expansion wave are the key factors affecting the change in the peak pressure inside the fully enclosed noise barrier, while the pressure of the semi-enclosed noise barrier is directly affected by slipstream of the train. (2) The pressure amplitude in the middle of the fully enclosed noise barrier is 2.1–2.7 times of that at the two ends. The amplitude of transverse force at the two ends of semi-enclosed noise barrier is 1.2–1.4 times of that in the middle. (3) In the longitudinal direction of the noise barrier near the train side, the pressure of the measuring point of the semi-enclosed noise barrier decreases with the increase in the measuring point height; The pressure in the middle section of the fully enclosed noise barrier is unaffected by the height of the measuring point. (4) A triangular buffer structure can effectively alleviate the unreasonable vertical pressure distribution phenomenon caused by the train bursting into the semi-closed noise barrier, and the reduction rate of the transverse force amplitude of the end noise barrier is as high as 21.44%. (5) The pressure relief scheme of three holes can effectively alleviate the phenomenon of excessive pressure amplitude in the middle of the fully enclosed noise barrier caused by the passing train, and the reduction rate of pressure amplitude can reach up to 60.5%. The pressure relief holes with an area of 1.7 H × 1.7 H are arranged at 0.25, 0.5, and 0.75 L, where L is the longitudinal length of the noise barrier.

Hydrodynamics of gas-liquid bubble columns under bubbling, transient, and jetting flow regimes using volume of fluid computational fluid dynamics

A volume of fluid computational fluid dynamics (VOF-CFD) model was used to investigate hydrodynamics of air-water bubble columns in the gas distributor region under bubbling, transient, and jetting flow regimes. The Reynolds-averaged Navier-Stokes (RANS) and large eddy simulation (LES) turbulence equations were coupled with the VOF-CFisms@2015D model. A novel bubble detection algorithm was developed to calculate the bubble size, where a gas fraction of 0.5 was adopted to detect bubbles in the gas-liquid mixture. The CFD results for all three flow regimes were compared to the experimental data in terms of the bubble size distribution (BSD) and mean bubble size. In the framework of VOF, the LES turbulence model captured bubble behaviors in the three flow regimes more accurately than the RANS approach by taking computational time 20% more. Increasing the surface tension from 0.025 to 0.0719 N/m resulted in the increment of the Sauter mean diameter from 4.0 to 5.4 mm. The VOF-CFD model coupled with LES yielded the initial BSD near the gas distributor, which is necessary for Eulerian CFD often applied to an entire column.

Numerical Modeling of Transient Two-Phase Flow and the Coalescence and Breakup of Bubbles in a Continuous Casting Mold.

The multiphase flow and spatial distribution of bubbles inside a continuous casting (CC) mold is a popular research issue due to its direct impact on the quality of the CC slab. The behavior of bubbles in the mold, and how they coalesce and break apart, have an important influence on the flow pattern and entrapment of bubbles. However, due to the limitations of experiments and measurement methods, it is impossible to directly observe the multiphase flow and bubble distribution during the CC process. Thus, a three-dimensional mathematical model which combined the large eddy simulation (LES) turbulent model, VOF multiphase model, and discrete phase model (DPM) was developed to study the transient two-phase flow and spatial distribution of bubbles in a continuous casting mold. The interaction between the liquid and bubbles and the coalescence, bounce, and breakup of bubbles were considered. The measured meniscus speed and bubble diameter were in good agreement with the measured results. The meniscus speed increased first and then decreased from the nozzle to the narrow face, with a maximum value of 0.07 m/s, and appeared at 1/4 the width of the mold. The current mathematical model successfully predicted the transient asymmetric two-phase flow and completely reproduced the coalescence, bounce, and breakup of bubbles in the mold. The breakup mainly occurred near the bottom of the submerged entry nozzle (SEN) due to the strong turbulent motion of the molten steel after hitting the bottom of the SEN. The average bubble diameter was about 0.6 mm near the nozzle and gradually decreased to 0.05 mm from the nozzle to the narrow face. The larger bubbles floated up near the SEN due to the effect of their greater buoyancy, while the small bubbles were distributed discretely in the entire mold with the action of the molten steel jet. Overall, the bubbles were distributed in a fan shape. The largest concentration of bubbles was in the lower part of the SEN and the upper edge of the SEN outlet.

Fast prediction of aerodynamic noise induced by the flow around a cylinder based on deep neural network

Accurate and fast prediction of aerodynamic noise has always been a research hotspot in fluid mechanics and aeroacoustics. The conventional prediction methods based on numerical simulation often demand huge computational resources, which are difficult to balance between accuracy and efficiency. Here, we present a data-driven deep neural network (DNN) method to realize fast aerodynamic noise prediction while maintaining accuracy. The proposed deep learning method can predict the spatial distributions of aerodynamic noise information under different working conditions. Based on the large eddy simulation turbulence model and the Ffowcs Williams–Hawkings acoustic analogy theory, a dataset composed of 1216 samples is established. With reference to the deep learning method, a DNN framework is proposed to map the relationship between spatial coordinates, inlet velocity and overall sound pressure level. The root-mean-square-errors of prediction are below 0.82 dB in the test dataset, and the directivity of aerodynamic noise predicted by the DNN framework are basically consistent with the numerical simulation. This work paves a novel way for fast prediction of aerodynamic noise with high accuracy and has application potential in acoustic field prediction.

Numerical Investigation of Plasma Actuator Effects on Flow Control Over a Three-Dimensional Airfoil With a Sinusoidal Leading Edge

Abstract The impetus of the present bio-inspired work is to investigate the impact of simultaneously using wavy leading-edge (WLE) airfoils in combination with curved multidielectric barrier discharge (DBD) plasma actuators as hybrid passive and active flow control mechanisms, respectively. A precise distinction of the produced frequency and noise signals, altogether with the acoustic effect of using WLE and DBD plasma actuators, is herein analyzed with precision. Two specific DBD plasma actuators are designed to actuate at x/C = 3% and x/C = 30% on a NACA 634-021 airfoil with sinusoidal WLE that bears a wavelength of 25% and an amplitude of 5% of the mean chord length and straight-leading-edge (SLE). A large eddy simulation (LES) turbulence model was used. This includes the dynamic control of unsteady flow separation, the three-dimensional vortical structure and induced trains of vortices, the aerodynamic forces, the velocity variation, and also the spanwise flow. The momentum transfer between the main flow and boundary layer was improved by the DBDs-induced vortices train and formed streamwise counter-rotating pair-of-vortices over the tubercle. Also, both the continuous wavelet transform (CWT) and fast Fourier transform (FFT) methods were used to investigate the induced plasma flow spectral content for the WLE and SLE geometries. We witnessed an optimized flow control, by using DBD plasma actuators with the WLE airfoil, that resulted in less massive flow separation, faster turbulent transition, and a robust earlier flow reattachment. This modification was beneficial in increasing the efficiency and decreasing the noise for low Reynolds number operational conditions.

Simulation of simplified frigate using embedded LES turbulence model: Part-I validation and turbulence scales

Understanding and characterization of turbulence over the flight deck of large marine vehicle is important to assess its influence on the helicopter pilot workload while performing launch and recovery maneuvers. This research is regarding understanding the turbulent flow field over the flight deck of simplified frigate model (SFS2) using an Embedded Large Eddy Simulation turbulence model in which a portion of flow field that is of interest is resolved using LES model, thereby reducing the overall computational cost. The influence of placement of RANS-LES interface is explored and detailed validation of the numerical scheme suggests that it can capture the turbulent flow structure, especially bi-stable air wake over the flight deck of SFS2 at the zero-yaw flow condition, with better accuracy. The development of downstream air wake into the far field is also discussed. Once the numerical scheme is validated, the characterization of integral time and length scales of turbulent vortical structures in the superstructure air wake is carried out using single point correlations. It is observed that flow field at flight deck center is influenced by vortices shed from top of superstructure and funnel whereas its edges towards port and starboard sides are predominantly affected by superstructure edge vortices.

LES modelling of multiphase turbulent flows in bubble columns using an adaptive-mesh finite element method

Turbulence drives mixing in multiphase flow reactors such as bubble columns, which are widely used in chemical and bioprocess industry. The understanding of turbulence in these two-phase gas–liquid columns is critical to understanding the dynamics of these flows for better equipment design. The present work focuses on the finite-element computational fluid dynamics modelling of turbulent flows in bubble columns with the use of the large eddy simulation (LES) model, which allows for resolving the dynamics of turbulent flows such as the oscillating plume. Two-dimensional fixed- and (optimized) adaptive-mesh simulations were performed and the results are validated against published experiments. The global flow parameters, including time-averaged gas holdup and plume oscillation period, showed good agreement with experiments. At a superficial gas velocity of 0.6cms−1, a plume oscillation period of 6.2s was predicted with a validation error of 9.6% and a gas holdup of 1.4% with a validation error of 7.5%. The novelty of the work lies in the use of a finite element framework with adaptive mesh refinement and the LES turbulence model to simulate bubble column reactors.

A WCSPH two-phase mixture model for tsunami waves generated by granular landslides

An advanced weakly compressible smooth particle hydrodynamics (WCSPH) two-phase mixture model is developed in this study to simulate the motion of granular landslides as well as their resulting tsunami waves. The governing equations of fluid and granular phases are established based on the two-phase mixture theory. The granular phase is assumed to be a weakly compressible non-Newtonian fluid, and the regularized visco-inertial rheology model is adopted to calculate its effective viscosity. The water phase is considered as a weakly compressible Newtonian fluid, and the large eddy simulation (LES) turbulence model is adopted to describe the turbulence effect of the fluid. The interphase force controls the momentum exchange between fluid and granular phases. The artificial diffusion term and particle shifting technique are adopted to improve the accuracy and stability of numerical calculation. This developed numerical model was applied to simulate submarine and subaerial granular landslide-tsunamis, and the results were compared with corresponding experimental data. The results show that the current model can adequately represent the generation and propagation of tsunami waves and the motion and deformation of granular landslides. Compared with the traditional two-phase granular flow models, the current model has higher accuracy and wider application scenarios.

Numerical studies on modeling the near- and far-field wake vortex of a quadrotor in forward flight

The operation of Unmanned Aerial Systems (UAS) is expected to keep increasing in the future with its rapid advancement and ever-expanding applications. One of the safety measures that need to be put into place for safe UAS operations is separation requirements. This article presents the results from flow simulations of quadrotor propellers using the Overset method in ANSYS Fluent 19.2 and the Virtual Blade Model (VBM) method in OpenFOAM as part of ongoing work to establish a wake vortex–based safe separation distance for UAS operations. The near-field flow simulation using the Reynolds-averaged Navier–Stokes turbulence model was validated with mesh convergence and turbulence model studies and verified with simulation and experimental data in the literature to ensure that the propeller simulation methods could generate a realistic near-field flow of multi-rotor UAS in forward flight. The far-field flow simulation using the Large Eddy Simulation turbulence model was validated with a mesh convergence study by calculating and comparing the circulation strength of the vortices in the far-field flow of each mesh setup but not verified due to a lack of existing experimental data.

CFD investigation of shock boundary layer interaction in hypersonic flow and flow control using micro ramps

PurposeThis study aims to investigate the prevalence of shock boundary layer interaction (SBLI) in air-breathing intake system is highly undesirable since this leads to high pressure gradients, typical stream mutilation and pressure drop. A novel flow control mechanism is incorporated in this research holding an array configuration of passive flow control device (micro ramps [MR]) that is adapted to improve the boundary layer stability.Design/methodology/approachTwo geometric variants of the MR, namely, MR40 and MR80 is considered which reduce the pressure drop during SBLI. The incidence oblique shock wave angle of 34° is considered for the modelling. Large eddy simulation (LES) turbulence model was used with subgrid models of Wall modelled LES, Smagorinsky–Lilly to compute the unsteady effects of SBLI control using micro vortex generators. The unsteady results are compared with steady Reynold’s average Naviers–Stoke’s equation for calibrating the turbulence models.FindingsThe array configuration of MR80 reduces the pressure drop by 22% as compared with no ramp configuration and also reduces the flow distortion in hypersonic inlet. The most affected region of the MR is in the vicinity of center-line. Quantitative results prove that the upstream influence of the shock waves has been largely reduces by MR80 array configuration as compared to single MR80 pattern configuration. Different vortex structures found in the experiments was exclusively predicted using LES.Originality/valueThis paper substantiates the requirement of MR array configuration for transferring the momentum from free stream to the boundary layer and thereby energizing the boundary layer. This process of energization delays the flow separation in hypersonic flow.

Experimental validation of liquid metal MHD numerical models

This study is aimed at experimental validation of a numerical model of fully coupled electromagnetic and fluid dynamics processes. The experiment is a laboratory-scale system, where a rotating permanent magnet induces an intense flow and free surface deformation of liquid Galinstan in a rectangular vessel. The numerical model is two-phase, 3D, transient, turbulent, and is constructed using ElmerFEM, OpenFOAM and EOF-Library codes. Validation is based on optical imaging of free surface projections. Statistical data, such as the time-average and standard deviation of the free surface profile, are compared with experiments and simulations. Results show that a Large Eddy Simulation turbulence model gives a good agreement with experiments, whereas a Reynolds-Averaged Navier-Stokes k-ω SST model causes artificial damping of free surface deformation. Figs 8, Refs 9.

Simulation of wind profile in coastal areas by using a structural wind-resistant moving vehicle tester: CFD and experimental investigations

In this study, based on the differentially spaced flat plates (FPs) technology, the wind profile with coastal terrain characteristics was simulated by using Computer Fluid Dynamics (CFD) and the transiting test method (TTM). First, CFD was adopted to establish the wind profile calculation models of the 2D and 3D TTM, and the computational domain, mesh types, boundary conditions, and other factors were determined. Second, the shear stress transport k–ω (SST k–ω) and large eddy simulation (LES) turbulence models were used to study the simulation results of the different interval between the FPs and the measuring points. Third, the suitable test conditions were selected from the numerical simulation results to guide the test conditions of the TTM. Lastly, the wind profile simulation results of the two methods were compared and analyzed to obtain the wind profile with coastal terrain characteristics. Results show that the wind and the turbulence intensity profiles obtained by the two methods are in good agreement with the GB 50009-2012 profile curve of exposure category A (i.e., coastal and other areas) when six FPs are used, the interval between the FPs and the measuring points is 240 mm–300 mm, and the inclination angle of the three FPs at the bottom is 7°.

The study on the wake of a CLT propeller under different advance coefficients

In this paper, the impact of advance coefficients on the characteristics of the CLT propeller wake is investigated and the large eddy simulation turbulence model is employed. A wide range of advance coefficients (0.1∼0.9) are calculated by solving the unsteady Navier-Stokes equations. The verification is performed and the results from the calculations are compared with experimental data. Under smaller advance coefficients, the wake is separated into three flow regimes: free stream, slipstream, and vortices between two regions. The boundaries among the flow regimes are indistinct under bigger advance coefficients. The perturbations induced by the rotating blades get smaller as the advance coefficient increases. Then, the vortical structures represented by the Q isosurface show that the endplate tip vortices merge with the secondary tip vortices, and the merged tip vortices will re-merge downstream. The inner trailing sheet vortices remain further than the outer ones under all advance coefficients. Finally, the trajectories (including the radius, the distance between two consecutive vortices, the pitch) of the vortex-system are analyzed in detail and fitted with different functions. It is recommended that proper considerations should be used for the establishment of the reliable wake model due to the various characteristics of the trajectory.

Analysis of Attack Angle Effect on Flow Characteristics Around Torpedo-Like Geometry Placed Near the Free-Surface via CFD

In this study, the flow characteristics of torpedo-like geometry placed near the free-surface at various angles of attack were investigated numerically. The study was carried out at the Reynolds number of Re=4x104 between immersion ratios of 0.75≤h/D≤3.5 and angles of attack α=0°,4°, 8°, and 12°. Large Eddy Simulation (LES) turbulence model was used along with the Volume of Fluid (VOF) multiphase model to investigate the effects of free-surface. Wake region had an asymmetrical structure near the free-surface as a result of the interaction. A jet-like flow region was observed between the geometry and the free-surface at lower immersion ratios due to the restriction of the flow area. This flow region had a downward movement towards the lower pressure wake region. The drag coefficient, CD, values were increased with the decrease of immersion ratio. At angles of attack α=8° and 12°, the flow separation occurring near the nose caused an additional restriction in the flow area and directed the jet-like flow toward free-surface. Variation of Froude numbers (Fr) depending on the immersion ratio is examined, and it is found that Fr number and corresponding drag coefficient have higher values for the lower immersion ratio. The free-surface effect was found negligible at h/D≥2.5 for all cases.

Improving Regularization Techniques for Incompressible Fluid Flows via Defect Correction

Abstract We propose and investigate two regularization models for fluid flows at higher Reynolds numbers. Both models are based on the reduced ADM regularization (RADM). One model, which we call DC-RADM (deferred correction for reduced approximate deconvolution model), aims to improve the temporal accuracy of the RADM. The second model, denoted by RADC (reduced approximate deconvolution with correction), is created with a more systemic approach. We treat the RADM regularization as a defect in approximating the true solution of the Navier–Stokes equations (NSE) and then correct for this defect, using the defect correction algorithm. Thus, the resulting RADC model can be viewed as a first member of the class that we call “LESC-reduced”, where one starts with a regularization that resembles a Large Eddy Simulation turbulence model and then improves it with a defect correction technique. Both models are investigated theoretically and numerically, and the RADC is shown to outperform the DC-RADM model both in terms of convergence rates and in terms of the quality of the produced solution.

Wind-interference effect on an octagonal building in varying positions by large eddy simulation

Three square plan shaped tall buildings are arranged at the vertices of an equilateral triangle of varying size. The octagonal plan shaped principal building is also positioned randomly such that the position of the principal building is never overlapped with the square plan shaped buildings. The effect of interference of the square buildings on the wind-induced responses of the principal building is thoroughly measured by computational fluid dynamics using large eddy simulation turbulence model. Metamodels for drag and lift coefficients of the principal building in the interfering conditions are established. Interference factors for the pressure coefficients at different faces of the principal building is calculated and cast into efficient mathematical models. Pareto optimality for the maximum interference criterion is obtained. The optimal positions of the principal and interfering buildings for maximum interference are plotted. Selected cases from the Pareto optima are analysed to obtain the transient wind effects for those interference cases. The peak and root-mean-square pressure coefficients of these interfering conditions are compared with that of the isolated condition. The difference between the power spectra of the isolated and interfering conditions are also depicted.

Numerical and Experimental Study of Inverse Diffusion LPG-Air Flames Pulsation

This study uses Ansys 16 commercial package to investigate an accurate numerical model that can trace the flame shape from inverse diffusion combustion of LPG with a focus on the effect of air pulsation on the combustion characteristics. The simulation is based on solving the energy, mass and momentum equations. The large eddy simulation turbulence model and the non-premixed combustion model are used to simulate the pulsating combustion reaction flows in a cylindrical chamber with an air frequency of 10,20,50,100 and 200 rad/sec. The numerical results are in great agreement with the experimental results in the flame shape and the temperature distribution along the combustion chamber in both pulsating and non-pulsating combustion. Diffusion combustion responds positively to pulsating combustion and increases mixing in the reaction zone. Increasing the air frequency increases the temperature fluctuations, the peak turbulent kinetic energy and maximum velocity magnitude, respectively, by 27.3%, 300%, and 200%. Increasing the Strouhal number to 0.23 shortens the flame by 40% and reduces nitric oxide and carbon monoxide by 12% and 40%, respectively, including an environmentally friendly combustion product. The maximum average temperature dropped from 1800 K to 1582 K with a very homogeneous temperature distribution along the combustion chamber which is very important for furnaces.

Numerical investigation of the effect of bluff body bluntness factor on the wake-vortex pattern

The research is concerned with the flow pattern and vortex shedding around a blunt trailing edge bluff body. The aim is to investigate the effect of the bluntness factor of the trailing edge on the fluid-flow for various aerodynamic purposes. This study is performed by using the blunt trailing edge plates with various thickness values ranging from 0.025 mm to 60 mm. Large Eddy Simulation (LES) turbulence model from ANSYS Fluent software is used to carry out the numerical simulation. The results show that the vortex formation range for the smallest bluntness thickness are relatively smaller than that of the larger bluntness thickness ratio. Generally, the intensity of the vortex street becomes stronger as the thickness of the plate increases, except for the blunt thickness of 10 mm due to the effect of the vortex eye location and the formation length. The influence of the bluntness thickness of the trailing edge has been identified as an important parameter to avert the drag issue associated with the blunt bodies. The most important aspect of this article is that the current study has unravelled that the vortex shedding occurs at the ratio below the known critical value. Further investigations are needed to better understand the underlying physics associated with the vortex shedding from various bluntness thickness.