Deep Learning Techniques for Effective Prediction of Aerodynamic Properties of Elliptical Bluff Bodies

2021 ◽  
Author(s):  
W. M. U. Weerasekara ◽  
H. M. C. D. B. Gunarathna ◽  
W. A. K. P. Wanigasooriya ◽  
T. P. Miyanawala
Keyword(s):  
Neural Network ◽  
Reynolds Number ◽  
Deep Learning ◽  
Aerodynamic Force ◽  
Flow Behavior ◽  
Reynolds Numbers ◽  
Bluff Bodies ◽  
Flow Conditions ◽  
Force Coefficients ◽  
Drag And Lift

Abstract Predicting aerodynamic forces on bluff bodies remains to be a challenging task due to the unpredictable flow behavior, specifically at higher Reynolds numbers. Experimental approaches to determine aerodynamic coefficients could be costly and time consuming. In the meantime, use of numerical techniques could also require a considerable computational cost and time depending on complexity of the flow behavior. The research focusses on developing an effective deep learning technique to predict aerodynamic force coefficients acting on elliptical bluff bodies for a given aspect ratio and given flow condition. Collecting data for drag and lift coefficients of several aspect ratios for flow conditions starting from onset of vortex shredding to verge of subcritical region is conducted by an accurate full order model. The specified region will provide a transient flow behavior and thus lift coefficient will be represented in terms of root mean square value and drag coefficient in terms of a mean value. With variations in flow behavior and vortex shredding frequencies, it requires to select an appropriate turbulence model, optimum discretization of fluid domain and time step to obtain an accurate result. Flow simulations are conducted primarily using Unsteady Reynolds Averaged Navier-Stokes Equations (URANS) model and Detached Eddy Simulations (DES) model. Effectiveness in using different turbulence models for specified flow regimes are also explored in comparison to available experimental results. At lower Reynolds numbers, aerodynamic force coefficients for a specified body will only depend on Reynolds number. But after a certain specific Reynolds number, aerodynamic forces are dependent on the Mach number in addition to Reynolds number. Therefore, for higher Reynolds numbers, aerodynamic force coefficients are recorded for multiple Mach numbers with same Reynolds number and will be fed to the neural network. With the development of the machine learning and neural network modelling, many of the fields have nourished and created effective and efficient technologies to ease complex functions and activities. Our goal is to ease the complexity in the computational fluid dynamic field with a deep neural network tool created to predict drag and lift coefficient of elliptical bluff bodies for a given aspect ratio with an acceptable accuracy level. Researchers have developed deep neural network tools to predict various flow conditions and have succeeded with sufficient accuracy and a satisfying reduction of computational cost. In our proposed deep learning neural network, we have chosen to model the network with inputs as the geometry setup and the flow conditions with validated drag and lift coefficients. The model will extract the necessary flow features into filters with the convolution operation performed on the inputs. Our main directive is to create a deep learned neural network tool to predict the target values within an acceptable range of accuracy while minimizing the computation cost.

A Novel Deep Learning Method for the Predictions of Current Forces on Bluff Bodies

2018 ◽  
Author(s):  
T. P. Miyanawala ◽  
Rajeev K. Jaiman
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Bluff Body ◽  
The Body ◽  
Stochastic Gradient Descent ◽  
Approximation Procedure ◽  
Bluff Bodies ◽  
Force Coefficients ◽  
Fluid Forces ◽  
Body Shapes

Unsteady separated flow behind a bluff body causes fluctuating drag and transverse forces on the body, which is of great significance in many offshore and marine engineering applications. While physical experimental and computational techniques provide valuable physics insight, they are generally time-consuming and expensive for design space exploration and flow control of such practical scenarios. We present an efficient Convolutional Neural Network (CNN) based deep-learning technique to predict the unsteady fluid forces for different bluff body shapes. The discrete convolution process with a non-linear rectification is employed to approximate the mapping between the bluff-body shape and the fluid forces. The deep neural network is fed by the Euclidean distance function as the input and the target data generated by the full-order Navier-Stokes computations for primitive bluff body shapes. The convolutional networks are iteratively trained using a stochastic gradient descent method to predict the fluid force coefficients of different geometries and the results are compared with the full-order computations. We have extended this CNN-based technique to predict the variation of force coefficients with the Reynolds number as well. Within the error threshold, the predictions based on our deep convolutional network have a speed-up nearly three orders of magnitude compared to the full-order results and consumes an insignificant fraction of computational resources. The deep CNN-based model can predict the hydrodynamic coefficients required for the well-known Lighthill’s force decomposition in almost real time which is extremely advantageous for offshore applications. Overall, the proposed CNN-based approximation procedure has a profound impact on the parametric design of bluff bodies and the feedback control of separated flows.

Heat Transfer for High Aspect Ratio Rectangular Channels in a Stationary Serpentine Passage With Turbulated and Smooth Surfaces

2013 ◽  
Author(s):  
Matthew A. Smith ◽  
Randall M. Mathison ◽  
Michael G. Dunn
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Aspect Ratio ◽  
Flow Behavior ◽  
Reynolds Numbers ◽  
Hydraulic Diameter ◽  
Internal Cooling ◽  
Number Range ◽  
Aspect Ratios ◽  
Parallel Configuration

Heat transfer distributions are presented for a stationary three passage serpentine internal cooling channel for a range of engine representative Reynolds numbers. The spacing between the sidewalls of the serpentine passage is fixed and the aspect ratio (AR) is adjusted to 1:1, 1:2, and 1:6 by changing the distance between the top and bottom walls. Data are presented for aspect ratios of 1:1 and 1:6 for smooth passage walls and for aspect ratios of 1:1, 1:2, and 1:6 for passages with two surfaces turbulated. For the turbulated cases, turbulators skewed 45° to the flow are installed on the top and bottom walls. The square turbulators are arranged in an offset parallel configuration with a fixed rib pitch-to-height ratio (P/e) of 10 and a rib height-to-hydraulic diameter ratio (e/Dh) range of 0.100 to 0.058 for AR 1:1 to 1:6, respectively. The experiments span a Reynolds number range of 4,000 to 130,000 based on the passage hydraulic diameter. While this experiment utilizes a basic layout similar to previous research, it is the first to run an aspect ratio as large as 1:6, and it also pushes the Reynolds number to higher values than were previously available for the 1:2 aspect ratio. The results demonstrate that while the normalized Nusselt number for the AR 1:2 configuration changes linearly with Reynolds number up to 130,000, there is a significant change in flow behavior between Re = 25,000 and Re = 50,000 for the aspect ratio 1:6 case. This suggests that while it may be possible to interpolate between points for different flow conditions, each geometric configuration must be investigated independently. The results show the highest heat transfer and the greatest heat transfer enhancement are obtained with the AR 1:6 configuration due to greater secondary flow development for both the smooth and turbulated cases. This enhancement was particularly notable for the AR 1:6 case for Reynolds numbers at or above 50,000.

Local Swirl Chamber Heat Transfer and Flow Structure at Different Reynolds Numbers

1999 ◽  
Vol 122 (2) ◽  
pp. 375-385 ◽  
Author(s):  
C. R. Hedlund ◽  
P. M. Ligrani
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Flow Structure ◽  
Flow Behavior ◽  
Turbine Blades ◽  
Reynolds Numbers ◽  
Local Flow ◽  
Flux Boundary Condition ◽  
Nusselt Numbers ◽  
Swirl Chamber

Local flow behavior and heat transfer results are presented from two swirl chambers, which model passages used to cool the leading edges of turbine blades in gas turbine engines. Flow results are obtained in an isothermal swirl chamber. Surface Nusselt number distributions are measured in a second swirl chamber (with a constant wall heat flux boundary condition) using infrared thermography in conjunction with thermocouples, energy balances, and in situ calibration procedures. In both cases, Reynolds numbers Re based on inlet duct characteristics range from 6000 to about 20,000. Bulk helical flow is produced in each chamber by two inlets, which are tangent to the swirl chamber circumference. Important changes to local and globally averaged surface Nusselt numbers, instantaneous flow structure from flow visualizations, and distributions of static pressure, total pressure, and circumferential velocity are observed throughout the swirl chambers as the Reynolds number increases. Of particular importance are increases of local surface Nusselt numbers (as well as ones globally averaged over the entire swirl chamber surface) with increasing Reynolds number. These are tied to increased advection, as well as important changes to vortex characteristics near the concave surfaces of the swirl chambers. Higher Re also give larger axial components of velocity, and increased turning of the flow from each inlet, which gives Go¨rtler vortex pair trajectories greater skewness as they are advected downstream of each inlet. [S0889-504X(00)00502-X]

scholarly journals Local Swirl Chamber Heat Transfer and Flow Structure at Different Reynolds Numbers

1999 ◽  
Author(s):  
C. R. Hedlund ◽  
P. M. Ligrani
Keyword(s):  
Heat Transfer ◽  
Reynolds Number ◽  
Flow Structure ◽  
Flow Behavior ◽  
Turbine Blades ◽  
Reynolds Numbers ◽  
Local Flow ◽  
Flux Boundary Condition ◽  
Nusselt Numbers ◽  
Swirl Chamber

Local flow behavior and heat transfer results are presented from two swirl chambers, which model passages used to cool the leading edges of turbine blades in gas turbine engines. Flow results are obtained in an isothermal swirl chamber. Surface Nusselt number distributions are measured in a second swirl chamber (with a constant wall beat flux boundary condition) using infrared thermography, in conjunction with thermocouples, energy balances, and in situ calibration procedures. In both cases, Reynolds numbers Re based on inlet duct characteristics range from 6000 to about 20000. Bulk helical flow is produced in each chamber by two inlets which ore tangent to the swirl chamber circumference. Important changes to local and globally-averaged surface Nusselt numbers, instantaneous flow structure from flow visualizations, and distributions of static pressure, total pressure, and circumferential velocity are observed throughout the swirl chambers as the Reynolds number increases. Of particular importance are increases of local surface Nusselt numbers (as well as ones globally-averaged over the entire swirl chamber surface) with increasing Reynolds number. These are tiad to increased advection, as well as important changes to vortex characteristics near the concave surfaces of the swirl chambers. Higher Re also give larger axial components of velocity, and increased turning of the flow from each inlet, which gives Görtler vnrtex pair trajectories greater skewness as they are advected downstream of each inlet.

Supercritical Reynolds number simulation for two-dimensional flow over circular cylinders

1975 ◽  
Vol 70 (3) ◽  
pp. 529-542 ◽  
Author(s):  
Edmond Szechenyi
Keyword(s):  
Surface Roughness ◽  
Reynolds Number ◽  
Reynolds Numbers ◽  
Practical Significance ◽  
Circular Cylinders ◽  
Wind Tunnels ◽  
Bluff Bodies ◽  
Wide Range ◽  
Two Dimensional Flow ◽  
Different Diameters

In wind-tunnel tests on bluff bodies the Reynolds number is often limited to values that are very much smaller than those of the flows being simulated. In such cases the experiments may have no practical significance whatsoever since both the fluctuating and the steady aerodynamic phenomena can vary considerably with Reynolds number.This difficulty was encountered in an investigation of supercritical incompressible flow over cylinders, and an attempt at artificially increasing the Reynolds number by means of surface roughness was made. In order to evaluate this simulation technique, the influence of various grades of surface roughness on the aerodynamic forces acting on cylinders of different diameters was studied over a wide range of Reynolds numbers in two very different wind tunnels. The results allow very positive conclusions to be drawn.

An approximate theory for the streaming motion past axisymmetric bodies at Reynolds numbers of from 1 to approximately 100

1977 ◽  
Vol 82 (3) ◽  
pp. 583-604 ◽  
Author(s):  
Michael S. Kolansky ◽  
Sheldon Weinbaum ◽  
Robert Pfeffer
Keyword(s):  
Boundary Layer ◽  
Reynolds Number ◽  
Pressure Gradient ◽  
Reynolds Numbers ◽  
Bluff Bodies ◽  
Approximate Theory ◽  
Viscous Term ◽  
Number Range ◽  
Curvature Effects ◽  
Flow Past

In Weinbaum et al. (1976) a simple new pressure hypothesis is derived which enables one to take account of the displacement interaction, the geometrical change in streamline radius of curvature and centrifugal effects in the thick viscous layers surrounding two-dimensional bluff bodies in the intermediate Reynolds number range O(1) < Re < O(102) using conventional Prandtl boundary-layer equations. The new pressure hypothesis states that the streamwise pressure gradient as a function of distance from the forward stagnation point on the displacement body is equal to the wall pressure gradient as a function of distance along the original body. This hypothesis is shown to be equivalent to stretching the streamwise body co-ordinate in conventional first-order boundary-layer theory. The present investigation shows that the same pressure hypothesis applies for the intermediate Reynolds number flow past axisymmetric bluff bodies except that the viscous term in the conventional axisymmetric boundary-layer equation must also be modified for transverse curvature effects O(δ) in the divergence of the stress tensor. The approximate solutions presented for the location of separation and the detailed surface pressure and vorticity distribution for the flow past spheres, spheroids and paraboloids of revolution at various Reynolds numbers in the range O(1) < Re < O(102) are in good agreement with available numerical Navier–Stokes solutions.

Unsteady flow about a sphere at low to moderate Reynolds number. Part 1. Oscillatory motion

1994 ◽  
Vol 277 ◽  
pp. 347-379 ◽  
Author(s):  
Eugene J. Chang ◽  
Martin R. Maxey
Keyword(s):  
Reynolds Number ◽  
Unsteady Flow ◽  
Steady Flow ◽  
Oscillatory Flow ◽  
Reynolds Numbers ◽  
Oscillatory Motion ◽  
Rigid Sphere ◽  
Flow Conditions ◽  
Steady Streaming ◽  
Moderate Reynolds Number

A direct numerical simulation, based on spectral methods, has been used to compute the time-dependent, axisymmetric viscous flow past a rigid sphere. An investigation has been made for oscillatory flow about a zero mean for different Reynolds numbers and frequencies. The simulation has been verified for steady flow conditions, and for unsteady flow there is excellent agreement with Stokes flow theory at very low Reynolds numbers. At moderate Reynolds numbers, around 20, there is good general agreement with available experimental data for oscillatory motion. Under steady flow conditions no separation occurs at Reynolds number below 20; however in an oscillatory flow a separation bubble forms on the decelerating portion of each cycle at Reynolds numbers well below this. As the flow accelerates again the bubble detaches and decays, while the formation of a new bubble is inhibited till the flow again decelerates. Steady streaming, observed for high frequencies, is also observed at low frequencies due to the flow separation. The contribution of the pressure to the resultant force on the sphere includes a component that is well described by the usual added-mass term even when there is separation. In a companion paper the flow characteristics for constant acceleration or deceleration are reported.

The effect of Reynolds numbers on the steady state aerodynamic force coefficients of the Stonecutters Bridge deck section

2008 ◽  
Vol 11 (3) ◽  
pp. 179-192 ◽  
Author(s):  
M.C.H. Hui ◽  
Z.Y. Zhou ◽  
A.R. Chen ◽  
H.F. Xiang
Keyword(s):  
Steady State ◽  
Bridge Deck ◽  
Aerodynamic Force ◽  
Reynolds Numbers ◽  
Force Coefficients

CFD Study of the Pressure Distribution on the Back of a 3-D Bluff Body (CUBE) of Air Flow in Different Reynolds Numbers

2009 ◽  
Author(s):  
Mohammad Javad Izadi ◽  
Pegah Asghari ◽  
Malihe Kamkar Delakeh
Keyword(s):  
Reynolds Number ◽  
Bluff Body ◽  
Free Stream ◽  
Fluid Flows ◽  
Reynolds Numbers ◽  
The Body ◽  
Stream Velocity ◽  
Bluff Bodies ◽  
Unsteady Forces ◽  
Flow Round

The study of flow around bluff bodies is important, and has many applications in industry. Up to now, a few numerical studies have been done in this field. In this research a turbulent unsteady flow round a cube is simulated numerically. The LES method is used to simulate the turbulent flow around the cube since this method is more accurate to model time-depended flows than other numerical methods. When the air as an ideal fluid flows over the cube, flow separate from the back of the body and unsteady vortices appears, causing a large wake behind the cube. The Near-Wake (wake close to the body) plays an important role in determining the steady and unsteady forces on the body. In this study, to see the effect of the free stream velocity on the surface pressure behind the body, the Reynolds number is varied from one to four million and the pressure on the back of the cube is calculated numerically. From the results of this study, it can be seen that as the velocity or the Reynolds number increased, the pressure on the surface behind the cube decreased, but the rate of this decrease, increased as the free stream flow velocity increased. For high free stream velocities the base pressure did not change as much and therefore the base drag coefficient stayed constant (around 1.0).

Numerical Prediction of Turbulent Flow in Rotating Cavities

1988 ◽  
Vol 110 (2) ◽  
pp. 202-211 ◽  
Author(s):  
A. P. Morse
Keyword(s):  
Reynolds Number ◽  
Turbulent Flow ◽  
Predictive Accuracy ◽  
Turbulent Transport ◽  
Reynolds Numbers ◽  
Spreading Rate ◽  
Flow Conditions ◽  
Wide Range ◽  
Ekman Layers ◽  
Radial Outflow

Predictions of the isothermal, incompressible flow in the cavity formed between two corotating plane disks and a peripheral shroud have been obtained using an elliptic calculation procedure and a low turbulence Reynolds number k–ε model for the estimation of turbulent transport. Both radial inflow and outflow are investigated for a wide range of flow conditions involving rotational Reynolds numbers up to ∼106. Although predictive accuracy is generally good, the computed flow in the Ekman layers for radial outflow often displays a retarded spreading rate and a tendency to laminarize under conditions that are known from experiment to produce turbulent flow.

Sign in / Sign up

Export Citation Format

Share Document