Internal flow is a very important topic in fluid mechanics and engineering. Many systems depend on the movement of liquids and gases inside pipes and closed channels. When we use CFD analysis of flow through a pipe, we can study the flow before building anything in the real world. This is important for water systems, oil and gas pipelines, heat exchangers, cooling channels, and many industrial machines.
Image: Pipe flow velocity field. Caption: CFD helps engineers understand the behaviour of internal pipe flow.
ANSYS Fluent is one of the most powerful tools for internal flow simulation. It can model pressure, velocity, turbulence, and heat transfer inside pipes with high accuracy. For many engineers, Fluent is the main software used for pipe flow CFD, turbulent pipe flow modeling, and heat‑transfer simulation in pipes.
This tutorial is a simple and complete step‑by‑step guide for simulating internal flow in a pipe in ANSYS Fluent. You will learn how to create the pipe geometry, build a good mesh, set the boundary conditions, run the solver, and study the results. All steps use short sentences and simple words so beginners can follow easily.
Figure 1: Wide variety of internal flow examples in numerical simulation in ANSYS Fluent software
To help with calculations, you can also use useful tools such as the Reynolds number calculator and hydraulic diameter calculator from CFDLAND:
These tools are helpful when starting any internal flow project.
By following this guide, you will:
- Learn how to run a CFD simulation of internal pipe flow.
• Understand each step from geometry design to final results.
This introduction prepares you for the full internal flow tutorial and helps you understand the basic ideas behind pipe flow simulation, pressure drop prediction, velocity distribution, and turbulent internal flow in ANSYS Fluent.
Initial Information
This article presents an internal flow simulation of water inside a circular pipe using ANSYS Fluent. The goal of the simulation is to analyze how heat and fluid move inside the pipe when the wall is heated with a constant heat flux. The pipe flow is turbulent, steady, and uses the incompressible flow model. The energy equation is activated for thermal analysis. The standard k‑epsilon turbulence model is selected for solving turbulence. The fluid enters the pipe at 298 K and heats up as it flows. This setup is common in many industrial systems, such as heat exchangers and cooling loops.
To create a realistic entrance flow, this simulation uses a developed velocity profile at the inlet. This profile is generated by first solving a pipe flow case in Fluent, then saving the outlet velocity and turbulence fields from that run as a profile file. This file is used as the inlet boundary condition in the main simulation. With this method, the flow enters in fully developed form. This removes the need for adding extra pipe length at the inlet and allows for a more stable and realistic velocity field from the start. It also defines the turbulent kinetic energy and turbulent dissipation rate correctly using data from the same profile file. The boundary conditions are standard: a velocity inlet with profile data, a pressure outlet, and pipe walls with constant heat flux. A detailed setup is provided in the boundary conditions section later.
One of the main simulation objectives is to calculate the Nusselt number along the pipe. This helps describe the convective heat transfer in the system. The value of the CFD-predicted Nusselt number is compared to the analytical value from the Dittus–Boelter correlation. Using the wall heat flux, wall temperature, bulk fluid temperature, Prandtl number, and Reynolds number, the heat transfer coefficient and Nusselt number are calculated. The result is validated by comparing the Fluent result to the value from theory. The final difference between the calculated Nusselt number and the Fluent result is only 1.63%, showing very good agreement. This confirms that the mesh, models, and solver setup are working correctly in this pipe flow simulation.
Geometry
In this step, we create the pipe geometry that we will use for the CFD simulation in ANSYS Fluent. The geometry is a simple and complete cylinder. It is created inside ANSYS DesignModeler.
Figure 2: Simple schematic of the pipe used for the CFD simulation.
The pipe has:
- Diameter: 0.05 meters
- Length: 1.5 meters
- Shape: Full cylinder (not half, not symmetry model)
- Software: ANSYS DesignModeler
These values are small and simple, so they help us focus on the main goal: simulating internal flow in a pipe.
Steps to Create the Cylinder in DesignModeler
- Open DesignModeler from ANSYS Workbench.
- Set Units to meters (m).
- Choose a plane (for example: XY plane).
- Create a 2D circle with a diameter of 0.05 m.
- Place the center of the circle at the origin for accuracy.
- Use the Extrude tool.
- Set the extrude length to 1.5 m.
- Generate the model to create the 3D cylinder.
This gives you a clean and simple pipe geometry for CFD analysis.
Figure 3: Full cylinder geometry (0.05 m diameter and 1.5 m length) created in ANSYS DesignModeler.
After creating the geometry, save your file. Now the pipe is ready for the meshing process.
Mesh
In this step, we create the mesh for the pipe in the ICEM software. A good mesh is very important for a correct CFD simulation of internal flow.
The pipe is a full cylinder with:
- Diameter: 0.05 m
- Length: 1.5 m
The mesh was made in ANSYS ICEM CFD, which is good for creating clean and high‑quality structured meshes. The mesh uses Hexahedral (Hex) cells. Hex cells are very good for pipe flow because the flow is smooth and long in one direction.
Table 1: Detailed mesh information based on ICM software
| Cells | Faces | Nodes |
| 1,081,195 | 3,268,152 | 1,106,400 |
This is a fine mesh, and it helps capture the velocity and pressure profiles inside the pipe. Mesh quality is very important. A good mesh makes the simulation stable.
From our ICEM mesh report:
- Minimum Orthogonal Quality: 0.7906
- Maximum Aspect Ratio: 10.50
These values show the mesh is clean and acceptable for pipe CFD simulation. The orthogonal quality plot shows that most cells have values close to 1.0, which is very good for internal flow.
Figure 4: Mesh size, mesh quality, and orthogonal quality report from ICEM.
Steps Used in ICEM
- The cylinder geometry was imported into ICEM.
- Blocking was created to match the pipe shape.
- Edge sizes were set to get fine cells across the diameter and length.
- Hex mesh was generated and checked for quality.
- The mesh passed ICEM quality checks and was exported for ANSYS Fluent.
This mesh is good for capturing the flow development and pressure drop in the pipe.
Boundary Conditions
In this simulation, the inlet and outlet boundary conditions are not typed by hand. Instead, a profile file is used. This file contains the velocity, turbulent kinetic energy (k), and turbulent dissipation rate (ε) at many points on the boundary. ANSYS Fluent reads these values and applies them to the inlet and outlet surfaces.
Inlet Boundary Condition
The inlet is set as a velocity‑inlet. The velocity is not a single number. It is a profile loaded from a file, using the point‑cloud method. The file gives Fluent a full extended velocity profile at the inlet. This means the inlet velocity follows the shape and distribution defined in the profile data, not a uniform value.
Table 2: Inlet Boundary Condition
| • Velocity direction | normal to boundary |
| • Velocity reference frame | absolute |
| • Velocity magnitude | taken from profile field velocity‑magnitude |
| • Turbulence model | k‑epsilon |
| • Turbulent kinetic energy (k) | from profile field turb‑kinetic‑energy |
| • Turbulent dissipation rate (ε) | from profile field turb‑diss‑rate |
| • Inlet temperature | 298 K |
| • Inlet pressure (initial guess only) | 0 Pa gauge |
| • Particle reinjection | keeps same injection assignment |
This lets the inlet match the exact flow distribution defined in the input file.
Outlet Boundary Condition
Table 3: Outlet Boundary Condition
| • Outlet boundary type | pressure‑outlet (or outflow, depending on the setup you choose) |
| • Velocity magnitude | from profile file |
| • k and ε | from profile file |
| • Pressure at outlet | 0 Pa gauge (default for pressure outlet) |
This allows the outlet to follow the same data‑driven distribution as shown in your profile window.
This CFD simulation uses a realistic velocity profile as the input at the pipe inlet. This profile was created by running a preliminary Fluent simulation of pipe flow. The outlet velocity field from this first simulation is then used as the inlet boundary condition for the main case.
This method means the flow enters with a fully developed shape. It is already expanded and realistic from the start. As a result, we do not need extra pipe length at the inlet to allow the velocity to develop naturally. This saves computational time and keeps the geometry short while keeping the momentum distribution accurate.
Figure 5: This profile file used, contains the velocity, turbulent kinetic energy (k), and turbulent dissipation rate (ε).
Wall Boundary Condition
The pipe wall uses no-slip motion, so the fluid at the wall has zero velocity. The wall material is aluminum, and the wall thermal condition is defined by heat flux. The applied heat flux is 50,000 W/m². Wall thickness and heat generation rate are both 0, and shell conduction is not enabled. The wall is stationary, and the convective augmentation factor is 1.
Figure 6: Boundary conditions related to applying a constant heat flux to the pipe wall in ANSYS Fluent software.
These boundary conditions complete the setup of how the flow enters, interacts with the pipe wall, and exits the pipe in the simulation.
Solver Settings
In this step, we set all the main options inside ANSYS Fluent. This includes physics models, materials, and important settings for internal flow simulation.
- Space: 3D
The pipe is a full 3D cylinder, so we use a 3D model. - Time: Steady
The flow does not change with time. We want a steady-state solution.
Figure 7: Solver settings related to solving internal pipe flow for this example.
- Viscous Model: Realizable k-epsilon
This model is good for internal pipe flow and gives stable results. - Wall Treatment: Enhanced Wall Treatment
This helps resolve the boundary layer near the pipe wall more accurately. - Heat Transfer: Enabled
Heat transfer is turned on because the material properties include thermal values.
Figure 8: Settings related to the viscosity model and energy model in ANSYS Fluent software.
We use two materials in the simulation:
Table 4: Tables related to the material of the pipe body and the fluid passing through the pipe.
Fluid: Water (liquid)
|
Solid: Aluminum (pipe wall material)
|
These material values are standard and suitable for heat‑enabled flow simulations.
Before running the simulation, we set the reference values. These values help Fluent calculate non‑dimensional numbers and monitor the flow. The reference area is set to 1 square meter, and the reference length is 1 meter. The fluid density is 1000 kg/m³, and the pressure reference is 0 pascal. The reference temperature is 298 K, and the reference velocity is 0.2 m/s. The viscosity used is 0.001 kg/(m·s), and the ratio of specific heats is 1.4. The Y‑plus value for heat transfer calculation is 300, and the reference zone is set to solid.
Figure 9: The values specified for the reference value correspond to the ANSYS Fluent settings for the solver.
The solver uses the pressure‑based method. The flow equation, turbulence equation, and energy equation are all activated because this simulation includes both turbulence and heat transfer. The absolute velocity formulation is turned on. Fluent uses under‑relaxation factors to control numerical stability.
The pressure factor is 0.3.
The density and body forces use a value of 1.
Momentum uses 0.7, and the turbulent kinetic energy and turbulent dissipation rate both use 0.8.
Turbulent viscosity is set to 1.
Figure 10: The Under Relaxation Factors for the ANSYS Fluent settings for the solver.
The solver uses the PISO method for pressure–velocity coupling. Skewness‑neighbour coupling is enabled. Both skewness correction and neighbour correction use a value of 1.
All equations use second‑order accuracy for better results. Pressure uses a second‑order scheme. Momentum, turbulent kinetic energy, turbulent dissipation rate, and energy all use second‑order upwind schemes.
Figure 11: Recommended settings for the Discretization Scheme for the numerical simulation in Ansys Fluent.
To prevent numerical errors, Fluent uses several safety limits. The minimum absolute pressure is 1 Pa, and the maximum absolute pressure is 5×10¹⁰ Pa. The minimum static temperature is 1 K, and the maximum static temperature is 5000 K. The minimum turbulent kinetic energy is 1×10⁻¹⁴, and the minimum dissipation rate is 1×10⁻²⁰. The maximum turbulent viscosity ratio is 100000.
After setting all values, we start the simulation. During the run, we watch the residuals. They must drop smoothly and reach a low value. We also check the velocity profile, pressure drop, and wall temperature to make sure the solution behaves correctly. When the residuals stop changing and the profiles look stable, the solution is considered converged.
Figure 12: Recommended settings for the calculation run to start the numerical simulation.
When the simulation converges, we can move forward to results and post‑processing.
Results
In this step, we look at the results of the CFD simulation.
We check the quality of the solution, the flow behaviour, and the heat‑transfer values inside the pipe.
Convergence Check
The simulation reached 646 iterations, and all equations fully converged.
The residual values for continuity, x‑velocity, y‑velocity, z‑velocity, energy, k, and epsilon all dropped below their limits.
This means the solver found a stable solution.
The residual plots show a smooth decrease. This is a scientific sign of a good solution because the equations become more balanced as the iterations continue.
Figure 13: The residual plots show a smooth decrease.
The monitored values also reached steady levels:
- The outlet velocity reached 0.21024 m/s
- The outlet temperature reached 305.17 K
- The heat‑transfer coefficient h reached 1098.399 W/(m²·K)
These values became flat in the graphs, which means the solution stopped changing. This is another strong sign of convergence.
Figure 14: Tables related to heat transfer coefficients (h) and average outlet velocity.
Heat‑Transfer Coefficient (h)
The average heat‑transfer coefficient h was calculated from the area‑weighted average of the wall. This value tells us how well the pipe wall transfers heat to the water. The value 1098 W/(m²·K) is reasonable for forced convection in a small, heated pipe. This agrees with scientific references for internal flow in pipes with turbulent regimes.
In this study, the pipe has heat applied to the wall. Because of the heating:
- The water warms up
- The density becomes slightly lower
- The velocity profile changes with temperature
This is called thermal expansion inside a pipe flow. It is important in many engineering systems like boilers, heat exchangers, and cooling lines.
The plots show that as the flow warms up, the temperature stabilizes and the outlet velocity reaches a constant value. This matches the behaviour expected for internal forced convection.
Important Internal Flow Parameters
This tutorial studied key ideas in internal pipe flow:
- Inlet and outlet velocity – These help define the flow regime and show how the flow develops along the pipe.
• Reynolds number – This number shows if the flow is laminar or turbulent. In this case, the flow is turbulent, so using the k‑epsilon model is correct.
• Pressure drop – Pressure drop along the pipe is an important factor in pipe design. It increases with pipe length, flow speed, and turbulence.
• Heat transfer and expansion – Heating the pipe wall changes the flow temperature and creates thermal expansion effects in the fluid.
Figure 15: Static pipe temperature contour at the end of the simulation in ANSYS Fluent software.
These concepts agree with standard CFD and fluid‑mechanics references that explain internal flow, turbulent heat transfer, and convection inside pipes. Overall, the simulation fully converged, the temperature and velocity stabilized, and the heat‑transfer coefficient was obtained accurately from the CFD model.
Validation
Validation is an important part of every CFD study. It helps us check that the numerical results from ANSYS Fluent are correct. In this pipe‑flow case, we use the Nusselt number to compare the Fluent results with the analytical value from heat‑transfer formulas. First, we calculate the Reynolds number (Re). The diameter of the pipe and the fluid properties are known. The relation is:
From the file: D = 0.05 m, U = 0.2 m/s, Re = 10000
Next, we calculate the turbulence intensity (T.I.) using the formula:
With Re = 10000, we get:
T.I. = 5.059644 %
In this simulation, a constant heat flux is applied to the pipe wall.
From the data: q” = 50000 W/m², T‑bulk = 302.799 K, T‑wall = 342.5189 K, k (thermal conductivity of water) = 0.6 W/m·K, h (convection coefficient from Fluent) = 966.63324 W/m²·K
To validate the simulation, we compare two Nusselt numbers:
- Nu (Fluent) – from the CFD result
- Nu (Analytical) – from heat‑transfer formulas
For heating of liquids in smooth circular pipes, the file uses the Dittus–Boelter correlation:
We have: Re = 10000, Pr = 6.97 (water at this temperature)
From the file: Nu (Analytical) = 79.253956
Fluent gives a numerical Nusselt number based on simulated wall heat flux and fluid temperature.
Nu (Fluent) = 80.55277
To validate the solution, we calculate the error between the two Nusselt values:
The final error is about 1.63%, which is very small. This means the simulation agrees well with the analytical solution and is valid. This confirms that:
- The mesh quality is good
• The boundary conditions are correct
• The heat‑transfer model is working correctly
• The CFD result is reliable
Conclusion
Internal flow simulation is important in fluid mechanics and engineering. It helps us study how fluids move inside pipes, channels, and tubes before building real systems. CFD reduces mistakes, saves time, and improves safety and efficiency.
Figure 16: CFD helps engineers see the flow inside pipes before building them
To get correct results in ANSYS Fluent, we must set the right input values. The key inputs are pipe geometry, fluid properties, inlet velocity, turbulence model, and mesh quality. Good mesh and correct boundary conditions give stable and accurate pipe‑flow CFD results.
The main output results are the pressure drop, velocity field, temperature field, turbulence values, and heat‑transfer coefficient. These results help engineers understand pumping power, heat‑transfer performance, and flow behaviour.
Figure 17: Pressure drop is an important result in internal flow CFD
Internal flow appears in many industries such as heating and cooling systems, chemical plants, and power plants. Engineers use CFD internal flow simulation to study these systems and improve designs. CFDLAND provides useful examples related to internal flow and heat transfer:
- Flow Through A Pipe CFD Simulation | Analytical Solution Validation
- Droplet Formation in Microfluidic Device CFD: A Fluent VOF Tutorial
- Shell-Side Performance CFD: A Fluent Validation for Shell and Tube Heat Exchangers
Figure 18: The CFDLAND examples of internal flow.
These examples show how turbulence, heat transfer, and pressure drop interact in complex geometries. They help students and engineers model internal flows correctly in Fluent. Scientific references such as the ASHRAE Handbook and the ANSYS Fluent User Guide support the importance of correct turbulence modeling, mesh setup, and boundary‑condition selection. Research on k‑epsilon and SST k‑omega models also improves predictions of turbulent pipe flow. Overall, this study shows that CFD is a strong tool for internal flow analysis, helping engineers solve real problems and improve fluid‑system design.