CFD Simulation of Flow Over a Cylinder Using ANSYS Fluent

🛠️ The Idea

For my first technical deep-dive post, I wanted to explore the fluid dynamics around a 2D cylinder, a classic benchmark problem in computational fluid dynamics. The goal was to simulate flow separation, vortex shedding and drag coefficient evolution, all of which make this an ideal starting point.

My aim was to:

1. Design the geometry and mesh it from scratch in ANSYS Workbench.

2. Configure boundary conditions and solver settings in ANSYS Fluent.

3. Extract engineering insights from velocity/pressure contours and drag coefficient graphs

🔧 Step 1: Geometry Creation

Using the DesignModeler module in ANSYS, I created a 2D rectangular domain with a centrally placed cylinder.

I ensured adequate spacing upstream and downstream of the cylinder to avoid boundary effects, using standard proportions.

🧩 Step 2: Meshing the Domain

I used the triangular meshing method, applying face sizing controls for refinement near the cylinder.

A structured mesh was generated with finer elements concentrated around the cylinder to accurately capture the boundary layer and wake development. The mesh size gradually increased as we moved away from the cylinder to optimize computational efficiency.

Simple refinement strategies were applied:

• A maximum element size was set near the cylinder surface to avoid losing detail.

• In downstream regions, uniform spacing (equally sized elements) was used to accurately resolve vortex shedding without bias.

I opted for a triangular mesh as a trade-off between resolution and simplicity for my first simulation.

🌊 Step 3: Defining the Physics in Fluent

I set up the simulation as 2D, steady, laminar flow using the pressure-based solver.

• Inlet: Velocity Inlet (1 m/s)

• Outlet: Pressure Outlet (0 Pa gauge)

• Cylinder & Walls: No-slip wall

• Working fluid: Custom-defined fluid with density = 1 kg/m³ and dynamic viscosity = 1 kg/m·s

Solver schemes used: SIMPLE for pressure-velocity coupling and second-order upwind for momentum equations.

Although the initial setup was steady, I later extended it to a transient simulation to capture the formation of the von Kármán vortex street - a repeating pattern of alternating vortices that emerge behind the cylinder at moderate Reynolds numbers. This allowed for a more realistic and dynamic visualisation of unsteady flow characteristics like fluctuating lift and drag forces.

📈 Step 4: Monitoring the Drag Coefficient

To ensure accuracy, I monitored the residuals for continuity and momentum equations. As seen below, the residuals decreased steadily and stabilised - confirming that the solution had converged numerically before extracting any force or flow data.

The results show the emergence of unsteady wake behaviour:

• The drag coefficient (Cd) initially spikes and then settles into mild oscillations, averaging around 1.2 - which is within a reasonable range for low Reynolds number laminar flow over a cylinder.

• The lift coefficient (Cl) displays clear, periodic oscillations - a classic indicator of alternating vortex shedding.

These patterns point to the formation of a von Kármán vortex street, even at Re = 80, highlighting how unsteady effects can still develop in idealised 2D laminar setups. The graphs reinforce the physical accuracy of the simulation and offer insight into bluff body aerodynamics in low-speed flow regimes.

🎨 Step 5: Post-Processing

This simulation helped me:

• Practice meshing strategy

• Understand solver settings and convergence behaviour

• Interpret drag/lift data and flow visualisation outputs

The animation clearly illustrates vortex shedding in the wake - known as the von Kármán vortex street. This visual cue aligns with the oscillations seen in the lift coefficient plot and enhances the physical understanding of unsteady flow behaviour in bluff bodies.

🔍 Reflections & What’s Next

This project solidified my fundamentals in CFD and gave me hands-on experience in handling solver stability, convergence, and interpreting time-dependent results.

Next, I aim to take on more advanced simulations involving:

• Turbulence modelling

• Multi-body interactions

• Heat transfer and conjugate heat flow

• Complex CAD geometries

💬 Let’s Connect

I’d love to hear feedback from fellow CFD enthusiasts, professors, engineers and students. If you’ve worked on similar simulations or have suggestions for interesting fluid flow cases, feel free to connect and drop a comment!