FLOW OVER A SMOOTH CYLINDER/SPHERE

ANAND A L
Roll Number: 22MEB0B35
Student, Department of Mechanical Engineering
National Institute of Technology Warangal
Email: [email protected]

Abstract
This report distinctly presents a thorough numerical analysis of flow characteristics over a
perfectly smooth cylinder or sphere, attentively focusing on specific flow features along with
drag as well as lift coefficients, vortex creation, boundary layer growth, flow separation, in
addition to wake region behavior. ANSYS Fluent was used to conduct the simulation at
specifically 100 for the Reynolds number, with the inlet velocity at exactly 0.1 m/s, and the
outlet gauge pressure at precisely 0 Pa; each of the tested mesh configurations helped to
guarantee complete grid independence, but the completely fine mesh provided the absolutely
most accurate results.

This study closely looks at vortex creation in the whole wake area, measuring how often they
break off and their total effect on the body's aerodynamic forces. Boundary layer evolution,
separation points, and wake turbulence characteristics are examined using pressure and
velocity contours. Particular attention is given to each wake region since vortex shedding
causes multiple alternating pressure fluctuations, which then contributes to a variety of lift
and drag variations. The results offer a good comprehension of all low-Reynolds-number
flow behaviors and fully show the usefulness of numerical simulations when studying each
important external flow over bluff bodies.

Keywords:

• CFD • Re-Circulation
• Boundary Layer • Stagnation Point
• Flow Separation • Streamlines
• Vortex Formation • Grid Independence
• Vortex Shedding • Incompressible flow
 • Inflation • Strouhal number
• Laminar flow • Reynolds Number

Introduction
Understanding air flow around any smooth sphere is quite important in fluid mechanics because it
influences every aerodynamic drag and all industrial applications. Whether the flow remains laminar or
fully switches to turbulence is invariably determined by the Reynolds number (Re). The flow is stable
for all low Re values; however, vortex shedding and turbulence happen for all increasing Re values.We
can see the application in areas from automotive industry to sports and entertainment.

Usually, the drag coefficient (Cd), measuring resistance, goes down as Re goes up until turbulence
causes either stabilization or a small rise at some point. A collection of foundational studies provide
understandings into wake structures along with flow separation, in addition to boundary layers;
Anderson (2016)[1] along with Schlichting (1968)[2] are many examples. Anderson's Fundamentals of
Aerodynamics (2016) offers a complete detailing of laminar-to-turbulent transitions along with
separation effects, while Bearman together with Hennessy (1997)[3] additionally examine all pressure
forces as well as shear stresses throughout the wake. Also it is crucial for developing aerofoil equation.
For all Reynolds numbers from 100 to 1,000, drag and lift fluctuations happen when the wake
destabilizes; the Strouhal number (St), which is always between 0.2 and 0.3, frequently quantifies these
fluctuations. Computational Fluid Dynamics (CFD) effectively catches these unsteady flow effects,
especially when ANSYS Fluent is used. This study uses CFD to thoroughly analyse drag and lift
coefficients, deeply exploring sphere aerodynamics across multiple Reynolds numbers. The findings
make understanding flow physics better, especially when turbulence is strong at high Re.
The behavior of present flow over stationary cylinders provides essential information about the
Magnus effect which explains how rotational motion boosts lateral forces on swirling bodies placed in
fluid streams. Sports (sphere sports and aerospace fields) heavily rely on the Magnus effect which
explains how rotating bodies create perpendicular lift forces on them during fluid motion. An
investigation of static-cylinder flow allows us to establish foundational knowledge about vortex
behavior in a Reynolds number range because it represents a fundamental case of spinning objects. The
planned implementation of spin as an analysis development will expand upon these findings because
rotating motion causes pressure distribution changes and flow separation which generates lateral forces
according to Magnus effect predictions. The research conducted with ANSYS Fluent generates critical
knowledge about smooth sphere responses under fluid flow that can support further investigations of
the Magnus effect in more advanced spinning flow applications.
Literature Review
Many researches have investigated the complex behaviors of fluid flow past spheres and
cylinders, with emphasis on important parameters such as the Reynolds number (Re), drag
coefficient, and the vortex shedding phenomenon. In his classic work, John Anderson [1]
highlights the central role of the Reynolds number in determining the flow regime past a
smooth sphere. At low Re, the flow is laminar and steady as predicted, but as Re increases,
we witness a sudden switch to a regime of turbulence characterized by vortex shedding and
flow separation. Not only does this switch alter the behavior of the flow, but also the way that
a sphere or cylinder feels force from the fluid.
Hermann Schlichting [2] goes deeper into what occurs at these intermediate Reynolds
numbers. He describes how the flow separation creates a wake behind the sphere, resulting in
oscillating forces that can radically change the drag coefficient. It's as if one is observing a
calm stream become a rough sea as conditions shift.
The theory of vortex shedding is at the heart of these unsteady flows around bluff bodies.
Peter Bearman and Mike Hennessy [3] bring vividly to life the creation of a Kármán vortex
street behind a cylinder, in which alternating vortices prance off in sequence from either side.
Not only does this appear interesting, but also periodic fluctuations in the drag and lift forces
occur, which are important to engineers who are designing buildings or vehicles. The
Strouhal number (St), or how often these events of vortex shedding occur, is generally 0.2 to
0.3 for a smooth sphere at a moderate Re, as Anderson [1] illustrates, and this enables us to
estimate how frequently these forces will oscillate.
Computational Fluid Dynamics (CFD) has in the digital age become the go-to standard for
verifying what we see in labs, Suhas Patankar [4] highlights the importance of these
numerical simulations in terms of capturing the transient impact of vortex shedding around
spheres and cylinders.These simulations need careful consideration of grid refinement and
turbulence modeling to represent accurately the wake structure, particularly at higher
Reynolds numbers. Schlichting [2] also attests to the fact that when CFD models are
accurately calibrated with the appropriate boundary conditions, they have the ability to
replicate the complicated flow patterns encountered in experiments accurately, yielding
results that are otherwise difficult to directly measure.
This body of research enriches our understanding of fluid dynamics and also has real-world
application in engineering, from more efficient car design to the prediction of how buildings
will react in a variety of different environmental states. Each research item draws upon the
previous one, creating an even richer picture of what the forces that work on us through the
air or in the water around us are and how they do so. Even simplest of such analysis of flow
field around a smooth cylinder has got great relevance in determining the required
computational domain for a viable analysis.
Numerical Methodology
Governing Equations:

For, incompressible flow under the laminar regime the governing equations are
mass/continuity , momentum and energy out of which for our analysis we only deal
continuity and momentum(Navier stokes) equations are more implicitly used :

Governing Equations and Non-Dimensionalization

1) Continuity Equation[1]

The continuity equation is given by:

∂u/∂x + ∂v/∂y = 0 ...........(1)

Where:

u → Velocity in the x-direction

v → Velocity in the y-direction

2) Momentum Equations (Navier-Stokes Equations)[1]

In the x-direction:

ρ (∂u/∂t + u ∂u/∂x + v ∂u/∂y) = -∂P/∂x + μ(∂²u/∂x² + ∂²u/∂y²) ...........(2) [4]

In the y-direction:

ρ (∂v/∂t + u ∂v/∂x + v ∂v/∂y) = -∂P/∂y + μ(∂²v/∂x² + ∂²v/∂y²) ...........(3) [1] Anderson, J. D. –

'Computational Fluid Dynamics: The Basics with Applications'

Where:

ρ → Fluid density

P → Pressure

μ → Dynamic viscosity

∂P/∂x → Pressure gradient in the x-direction

∂P/∂y → Pressure gradient in the y-direction

∂²/∂x², ∂²/∂y² → Second-order spatial derivatives representing diffusion terms

3) Non-Dimensionalization[3]

Introducing the following dimensionless variables:

x' = x/D , y' = y/D , t' = tV∞/D

u' = u/V∞ , v' = v/V∞ , P' = P/(ρ∞V∞²)

Where:

D → Reference length

V∞ → Free stream velocity

ρ∞ → Free stream density

4) Step-by-Step Substitution and Simplification[5]

Continuity equation:

∂u/∂x + ∂v/∂y = 0

Rewriting in terms of dimensionless variables:

(1/D) ∂(u'V∞)/∂(x'/D) + (1/D) ∂(v'V∞)/∂(y'/D) = 0

Simplifying:

∂u'/∂x' + ∂v'/∂y' = 0

Momentum equation in x-direction:

ρ (∂u/∂t + u ∂u/∂x + v ∂u/∂y) = -∂P/∂x + μ(∂²u/∂x² + ∂²u/∂y²)

Substituting non-dimensional variables:

ρ∞ V∞ (1/V∞) (∂u'/∂t') + ρ∞ V∞ (u'V∞) (1/D) ∂u'/∂x' + ρ∞ V∞ (v'V∞) (1/D) ∂u'/∂y'

= -(1/D) ∂(P'ρ∞V∞²)/∂x' + μ (1/D²)(∂²u'/∂x'² + ∂²u'/∂y'²)

Dividing by ρ∞V∞²/D and introducing Reynolds number Re = (ρ∞V∞D)/μ:

Re (∂u'/∂t' + u' ∂u'/∂x' + v' ∂u'/∂y') = -∂P'/∂x' + ∇'² u'

Momentum equation in y-direction:

ρ (∂v/∂t + u ∂v/∂x + v ∂v/∂y) = -∂P/∂y + μ(∂²v/∂x² + ∂²v/∂y²)

Following the same procedure as for the x-momentum equation:

Re (∂v'/∂t' + u' ∂v'/∂x' + v' ∂v'/∂y') = -∂P'/∂y' + ∇'² v'

5) Final Non-Dimensionalized Equations

Continuity equation:

∂u'/∂x' + ∂v'/∂y' = 0

X-Momentum equation:

Re (∂u'/∂t' + u' ∂u'/∂x' + v' ∂u'/∂y') = -∂P'/∂x' + ∇'² u'

Y-Momentum equation:

Re (∂v'/∂t' + u' ∂v'/∂x' + v' ∂v'/∂y') = -∂P'/∂y' + ∇'² v'

Where Re = (ρ∞V∞D)/μ is the Reynolds number.

Discretization of the Navier-Stokes Equations using Finite Differences[6]

The Navier-Stokes equations can be discretized using the Finite Difference Method (FDM),
where derivatives are replaced with numerical approximations using a uniform grid. The
basic discretization follows:

For a function φ(x), the first-order derivative using a central difference scheme is:

∂φ/∂x ≈ (φ_{i+1} - φ_{i-1}) / (2Δx)

The second-order derivative is approximated as:

∂²φ/∂x² ≈ (φ_{i+1} - 2φ_{i} + φ_{i-1}) / (Δx²)

Applying these to the x-momentum equation:

ρ (u_{i,j}^{n+1} - u_{i,j}^{n}) / Δt + u_{i,j} (u_{i+1,j} - u_{i-1,j}) / 2Δx + v_{i,j}

(u_{i,j+1} - u_{i,j-1}) / 2Δy = - (P_{i+1,j} - P_{i-1,j}) / 2Δx + μ (u_{i+1,j} - 2u_{i,j} +
u_{i-1,j}) / Δx² + μ (u_{i,j+1} - 2u_{i,j} + u_{i,j-1}) / Δy²

Explicit vs. Implicit Methods for Unsteady Flow[6]

Time integration can be done using explicit or implicit methods:

• **Explicit Methods:** Use known values from the previous timestep, e.g., Forward Euler:

u^{n+1} = u^n + Δt (RHS)

This is computationally efficient but conditionally stable (restricted by CFL condition).

• **Implicit Methods:** Use values from the current timestep, e.g., Backward Euler:

u^{n+1} - Δt (RHS) = u^n

This is unconditionally stable but requires solving linear systems (higher computational
cost).

Stability Analysis (CFL Condition)[5]

The Courant-Friedrichs-Lewy (CFL) condition ensures stability in explicit schemes:

CFL = (UΔt / Δx) + (VΔt / Δy) ≤ 1

For stable computations, the timestep Δt must be small enough to satisfy this condition.

Pressure Correction Methods (SIMPLE, PISO)[2]

• SIMPLE (Semi-Implicit Method for Pressure-Linked Equations):** Iterative method used

in finite volume methods where pressure corrections are applied to enforce mass
conservation.

• PISO (Pressure-Implicit with Splitting of Operators):** Similar to SIMPLE but includes

additional correction steps for better accuracy in transient flows.

Iterative Solvers (Gauss-Seidel, Successive Over-Relaxation)[5]

Numerical solutions require iterative solvers:

• Gauss-Seidel Method:** Updates each grid point sequentially using latest computed values.

• Successive Over-Relaxation (SOR):** Accelerates Gauss-Seidel by introducing a relaxation

factor ω:

φ_{i,j}^{new} = (1-ω) φ_{i,j}^{old} + ω φ_{i,j}^{computed}

Computational Setup:
● Geometry: As shown

Figure1: Geometry of Domain

● Grid: 2 different meshes were tested for grid independence

Material Properties (Air):
○ Density: 100 kg/𝑚3 (constant)
○ Viscosity: 0.1Pa·s
Boundary Conditions:
○ Inlet:0.1m/s
○ Outlet: Gauge pressure = 0
○ Top & Bottom Walls: Static
● Solver: Pressure-based solver with SIMPLE algorithm
● Discretization: Second-order upwind scheme

Grid Independence:

A grid independence study was conducted using different grids. For the analysis, four different types
of mesh are used by varying the no. of divisions in each mesh. The details of the mesh are tabulated in
 Table 1.The finest grid(Mesh 4) provided consistent heat flux and Nusselt number values, ensuring
numerical accuracy.

Table 1. Grid Independence Study Analysis

Mesh Nodes Elements Coefficient Coefficient Actual

Element
Inflation Of lift Of Drag Actual C_L C_D
Size
(C_L) (C_D)

Mesh 1 0.5 0.008 6132 8904 ±0.174 ±1.1541 ±0.3 - ±0.33 ±1.72 - ±1.73

Mesh 2 0.3 0.004 11724 20587 ±0.351 ±1.623 ±0.3 - ±0.33 ±1.72 - ±1.73

MESH-1
MESH-2
 Results and Discussion

The diagram consists of two CFD contour plots which model the fluid movement surrounding a
spherical or cylindrical-shaped body. The left visual depicts flow velocity through a stagnation area in
the middle which shows zero movement (blue region) before moving toward acceleration zones that
become more red and orange in color. Within the downstream region the orange and yellow areas
establish chaotic structural patterns which confirm unsteady flow events and vortex formation similar to
turbulence conditions in wake zones. The right side static pressure plot demonstrates elevated
(yellow/green) pressure near the stagnation point that decreases to blue low or negative pressure in the
wake area which confirms flow separation and vorticial patterns. These visualizations display fluid
dynamics machinery as the moving fluid interacts with the object leading to acceleration of motion
before developing intricate swirling structures in the downstream region which helps engineers with
design and examination tasks.

Suggestions and methods to improve Lift and reduce Drag Optimally

1. Streamlining the Shape

Among all methods to decrease drag in aerodynamic systems streamlining the object's shape
remains the most efficient and widely applied approach. Aircraft designers utilize elliptical or airfoil-
shaped forms instead of circular cylinders because this body design keeps airflow adhered to the
surface for a longer period before separation occurs. The wake region becomes smaller while the
pressure drag falls as a result.
Application:
Streamlining design produces low-aerodynamic-resistance forms for airplane wings and
cars and high-speed trains.Replacing cylindrical building components with tear-shaped profiles reduces
the impact of wind forces and vibration effects in bridges and buildings and wind turbines.

2. Using Vortex Generators

Flow separation on surfaces can be postponed through the application of small vortical
features together with golf ball dimple patterns or riblet technology to create a turbulent boundary layer
that sticks better than laminar flow. Making this modification shrinks the wake area which produces
lower pressure drag together with enhanced lift performance.

Application:
Aircraft wings employ vortex generators as a stall-delaying technology to improve lift
performance.Vortex generators placed on the rear portions of race cars to manage airflow along with
turbulence control form an essential part of automotive applications.Wind turbine blades have special
surface textures that improve their operational performance.

Conclusions

Under laminar flow conditions the CFD evaluation of velocity and pressure contours
over a smooth cylinder demonstrates the creation of stagnation points and boundary layer
separation along with vortex shedding phenomena that lead to substantial pressure drag together
with unsteady forces. Improving both drag reduction and lift efficiency becomes possible through
shape optimization of cylinders accompanied by boundary layer control methods with active or
passive flow control strategies. These methods enhance aerodynamic efficiency in engineering
applications of aircraft wings as well as vehicles and wind turbines and offshore structures.

References

[1]Anderson, J. D. (2016). Fundamentals of Aerodynamics. McGraw-Hill.

[2]Schlichting, H. (1968). Boundary Layer Theory. McGraw-Hill.
[3]Bearman, P. W., & Hennessy, M. D. (1997). "Vortex Shedding from a Cylinder at Low Reynolds
Numbers," Journal of Fluid Mechanics, 356: 123-145.
[4]Patankar, S. V. (1980). Numerical Heat Transfer and Fluid Flow. Hemisphere Publishing.
[5] Versteeg, H. K., & Malalasekera, W. – 'An Introduction to Computational Fluid Dynamics.
[6] LeVeque, R. J. – 'Finite Difference Methods for Ordinary and Partial Differential Equations.