CFD simulation

Aerotak is a fluid mechanics and Computational Fluid Dynamics CFD consultancy company. Offering CFD consultant, CFD specialist, fluid mechanics consultant, fluid mechanics specialist. We couple engineering development with fluid dynamics and CFD. Running design optimisation using CFD and fluid dynamics. We validate our CFD simulations against experimental test results.

Aerotak er en konsulentvirksomhed som tilbyder Computational Fluid Dynamics CFD konsulent og fluid mekanik konsulent. Aerotak har expertise inden for ANSYS Fluent, Siemens CCM+ og OpenFOAM. Aerotak har stor erfaring inden for procesindustrien og bilindustrien. Aerotak tilbyder fluid specialist, CFD specialist, samt fluid konsulent og CFD konsulent.

 

CFD simulation - What is computation fluid dynamics?

Computational fluid dynamic simulations is the use of numerical simulations to analyze fluid dynamics. Aerotak™ are specialist with CFD simulation and the following is a simple example shown who to conduct a CFD simulation.


What is a CFD simulation?

 

Computational fluid dynamics (CFD), is the simulation of a fluids dynamics and its interaction with the surroundings. Aerotak™ uses CFD simulation in the prediction of what happens when fluid flow. Fluid flow covers everything from breathing to how an airplane works. Using CFD simulations, Aerotak™ is able to qualitatively and quantitatively predict the movement of air through your lungs or evaluate the performance of an airplane for optimization and design purposes. By use of CFD simulations a broad range of flow phenomena can be solved for when in possession of the proper experience, knowledge and technical skills. These phenomena include:

  • Mass transfer (e.g. sediment transport, evaporation)

  • Heat transfer (e.g. ovens, combustion)

  • Chemical reactions (e.g. combustion, reaction kinetics)

  • Forces on solid bodies (e.g. cars, monopiles, airplanes)

As fluid dynamic challenges come in many forms the CFD simulation of these must be handled with care and focus on the goal of the simulation. The complexity, time and resources required to run the simulation is highly dependent on the goal of the simulation. To explain the process and highlight the importance of simulating with respect to the desired goal a short case is shown below.


Flow around a cylinder case study

For the purpose of this case study, we give ourself the challenge of understanding the fluid flow around a cylinder.

1. Define the goal of the simulation

The first and most important step is always to define the goal of the simulation. The purpose of the present simulation is to understand the flow around a cylinder, which is a very vague purpose. Therefore, we now set the specific goal of visualizing the transient velocity field generated by the cylinder.

2. Prepare the geometry

Having defined the problem statement, we need to define the geometry of the fluid domain and the operating conditions of the flow. To simplify the governing equations and reduce computational effort, the flow can be considered two-dimensional, thus the fluid flow is predominant in only two-dimensions, whereas the third direction (into the paper) is neglected and not solved for.
Boundary conditions, to specify the conditions of the flow, are imposed on the surface of the cylinder, as well as the faces of the bounding box of the fluid domain. This entails setting inlet velocity, no-slip conditions (zero velocity) on walls etc.

The fluid domain is indicated by the blue color in the figure below. Within the cylinder (white color) the flow is not solved for.

Computational Fluid Dynamic (CFD) geometry of a cylinder in a flow

3. Mesh

CFD and numerical methods enable us to solve for complex flows to which no analytical solution exists. In order to do so the equations governing the flow are transformed into a set of algebraic equations to be solved for on smaller subdomains over the entire fluid domain. The process of creating these subdomains (or cells) is referred to as spatial discretization of the fluid domain, or simply meshing. In general, the more cells the more accurate the solution with respect to the real flow. The increased accuracy comes at the cost of increased computational time. Thus, the area of interest and areas of largest local changes in the flow (e.g. close to the and downstream of the cylinder) are given the smallest cell sizes. Here, the two-dimensional flow assumption comes in handy, since the number of cells, and the size of the equation system is significantly reduced.

A discretized version of the fluid domain is shown below. Each square is the face of a three-dimensional cell.

Computational Fluid Dynamic (CFD) Mesh for flow around a cylinder

4. Solve

For a large range of fluid mechanics problems, the full governing Navier-Stokes equations need not be solved. Averaging the terms of the equations, yields a simplified set of equations, known as RANS (Reynolds Averaged Navier-Stokes) equations. The application of RANS requires addition of a so-called turbulence model. Turbulence models are required for simulation of almost all every-day phenomena and industrial processes, where the flow is said to be turbulent.  The choice and implementation of turbulence model has a significant influence on the outcome of the simulation as well as the computational time.

The flow around the cylinder is highly unstable and turbulent in most cases. Thus, besides a turbulence model, one needs to solve the governing equations for each mesh cell at a series of specific times, since the flow field changes with time. In addition to the spatial discretization, a temporal discretization is required for stepping forward in time.

A solution to the flow around the cylinder is visualized below at a specific point in time. The presence of the cylinder imposes induced velocities in the onset flow, creating areas of low and high velocities downstream (behind) of the cylinder.

Computational fluid mechanics is a time-consuming process when it comes to solving for the flow field. Computational times are highly dependent on mesh quality and type, time-step size, mesh size, convergence rates per time-step, programming language, hardware, and the implemented numerical algorithms. At Aerotak™ we only use state-of-the-art software and hardware for our simulations and we are specialists in creating optimized computational setups, to swiftly supply highly accurate results at low cost.

Velocity field for flow around a cylinder with computational Fluid Dynamic (CFD)

5. Post-process and visualize

In relation to the goal of the study, necessary information must be extracted for the simulation results. For the case of the flow around the cylinder, visualization of the flow field is simply a way of representing numbers as images, allowing for a qualitative investigation of the flow behavior. Other relevant post-processing steps could entail calculation of derived quantities (e.g. vorticity) and integral parameters (e.g. lift and drag on cylinder along with the flow oscillation frequency), as well as data analysis by means of statistical tools.

Velocity field of the flow around a cylinder obtained with Computational Fluid Dynamic (CFD)

6. Take it to the next level

The unsteady shredding of vortices in the wake of the cylinder can be visualized as a function of time. The vortices are convected downstream to form a ‘street’ of vortices or eddies, commonly known as a von Karman vortex street.

The prediction of the behavior of such a vortex shedding flow is important in many cases including mooring of monopiles, noise and vibration control for cables and wires, wind load analyses of towers and so forth. Thus, a simply study as the one conducted here, can provide valuable information to a broad range of physical problems.

Do you have a Computational Fluid Dynamics project, contact us here.