flare tip mesh diagram
6 min

Fluid Flow Simulations with Computational Fluid Dynamics CFD

Posted by: Dr Lefki Germanou Date: 31 May 2024

Produced by CFD Consultants: David Braybrook, Calum McLaughlin, Dr Lefki Germanou

Using CFD simulations for fluid flow and heat transfer analysis

Computational Fluid Dynamics (CFD) is a branch of fluid mechanics used to solve and analyse problems involving fluid flow and heat transfer using numerical methods and algorithms. CFD allows engineers and researchers to study complex fluid flow phenomena, predict the performance of fluid flow systems and optimise designs without the need for costly and time-consuming physical prototypes and testing.

CFD has developed immensely in the last decade as a result of big leaps in algorithmic research, code development and computing power. Whereas previously, CFD required the use of supercomputers to simulate even simple models, the development of technology has enabled the simulation of cases involving physics and geometries of far greater complexity. This has allowed engineers and scientists to achieve more accurate and detailed results, with faster turnaround times, in new emerging industries. Hence CFD has become a commonly used tool for solving real-world physics and engineering problems.

What is needed to perform CFD fluid flow simulations?

To perform CFD simulations we first need to define the problem to be solved. This includes specifying the geometry to be modelled, the phenomena to be simulated and the field values at the boundaries of the model (such as the temperatures, pressures, velocities, heat sources).

In the context of consultancy work, usually the CFD consultants discuss with the client their objectives from the CFD analysis and proceed with the problem definition and modelling approach. A CAD model or drawings are supplied, in order to create a suitable geometry and extract the fluid volume to be used in the model. Furthermore, the properties of the fluid(s), such as density and viscosity, need to be known, as well as the process conditions (e.g., imposed flowrates, temperatures). Once the above information is provided the CFD process can begin.

Geometry

The CFD modelling workflow usually starts with the creation of a suitable “watertight” geometry which typically comes from a CAD model or drawing. This geometry is then included in the computational domain, which is the space the fluid passes through. In the case of a simple flow in a straight pipe, the computational domain is the void inside the pipe.

Mesh

The partial differential equations that govern fluid flow and heat transfer are highly non-linear and hence must be discretised and solved numerically. Therefore, to analyse fluid flows, the computational domain is also discretised into smaller subdomains. The discretised governing equations are then solved inside each of these subdomains to calculate properties like velocity and pressure. The subdomains are often called elements or cells, and the collection of all elements is called a mesh. Figure 1 displays a surface mesh of a flare tip.

flare tip mesh diagram

Figure 1: Example of a flare tip surface mesh

 

Simulation

All CFD simulations require the use of models which best represent the physical problem to be solved. For the setup, we need to decide if we are interested in the evolution of the flow in time, therefore using a transient simulation, or aim for the steady-state solution of the flow. Moreover, we should ask ourselves if the flow is compressible or not, and if we need to account for the effect of turbulence, hence using appropriate turbulence models depending on the Reynolds number.

The complexity of the simulation can increase if the problem involves two or more phases, i.e., air and water or air and water and oil. These models are referred to as multiphase models and are more computationally expensive. Regardless of the type of problem, there must be careful consideration in the selection of models to correctly predict the fluid flow and represent the physics. Figure 2 shows an example of a combustion simulation at a flare tip.

flare tips combustion results diagram

Figure 2: Example of flare tip combustion simulation results

 

Analysis

During the CFD simulation, the governing equations are solved with an iterative process until the numerical solution is converged. This means that the solution starts from an initial guess and gradually evolves to the final, unchanged flow field (for steady-state simulations). Then the CFD results can be analysed. This may be a case of extracting the pressure drop across two points, for example across an orifice plate, or possibly something more complex such as the combustion efficiency of a flare tip. CFD offers significant flexibility in analysing results, which is mostly done post-simulation.

Equations and Models

The primary equations solved in all CFD codes are the Navier-Stokes equations, a set of partial differential equations that describe how fluid velocity, pressure, density and viscosity of the fluid evolve in space and time. These equations are non-linear and must be solved numerically.

However, for most real-world engineering flows, there is also the added complexity of turbulence. Turbulence remains one of the oldest unsolved problems in physics. Typically, in computational fluid dynamics, turbulence is modelled using turbulence models, which are mathematical formulations that approximate the effects of turbulence on fluid flow. These models aim to capture the statistical characteristics of turbulence without explicitly resolving all turbulent scales, which would be computationally prohibitive. Many different turbulence models of different levels of complexity exist. Each has its own benefits and drawbacks, and it is up to experts to choose a model that is most applicable to the problem they are trying to solve.

Another phenomenon that is sometimes encountered in fluid flow modelling is phase-change. This results in a multiphase flow containing liquid and gas due to phenomena such as vaporisation due to pressure drops or evaporation from heating. Again, different models such as the volume of fluid model or mixture model are available to approximate these flows, and other settings such as numerical methods and thermodynamic models need to be chosen to suit the problem. Similarly, different models are available for chemical reaction or erosion simulations.

Which industries benefit from CFD analysis?

CFD is most commonly associated with the aerospace and automotive industries. In the automotive sector, CFD is employed to optimise vehicle aerodynamics and passenger comfort, reduce noise, and enhance overall performance. Similarly, in the aerospace industry it is used in the design of aircraft ensuring optimal aerodynamic performance, stability and fuel efficiency.

However, CFD also serves as a vital tool in advancing innovation in other domains, such as the energy sector and chemical and process industries.

Transportation & Aerospace

One of the first successes of CFD in the aerospace industry came from the design of the fuselage, wings and engines of the Boeing 737. Its significance lay in the substantial cost and time savings over wind-tunnel testing, and impracticality of building scaled models. Recently though, CFD has become a powerful tool, used in many other areas such as aeroelasticity of aircraft wings, engine combustion, and aeroacoustics.

Chemical & Process Industries

CFD has demonstrated success in improving performance of process equipment such as heat exchangers, mixers and reactors, and instruments such as valves and flow meters. Flow metering applications commonly benefit from CFD analysis, which includes evaluating differential pressure flow meters and correcting flow velocity of ultrasonic meters.

CFD is particularly useful in this area as it minimises the need for oil & gas operators to physically modify, test or interact with the metering lines, minimising loss of production. Research work recently undertaken by TÜV SÜD with funding from DSIT, provides confidence in the use of CFD-based correction factors for ultrasonic flow meters used in flare gas measurement.

Energy Sector

With the drive to decarbonise the energy sector, CFD is not only integral to traditional chemical and process industries reliant on fossil fuels but is also increasingly applied in the low- and zero-carbon energy sectors. Several areas benefitting from the use of CFD include computational wind turbine blade design; wind farm wake analysis and mitigation of these disturbances; safe coolant system design, heat analysis and safety systems in nuclear power; and modelling fluid behaviour and metering in carbon capture, storage and utilisation (CCUS) systems. CFD’s value in this context lies in its facilitation of research and development before production, minimising costs and increasing efficiency.

Benefits of Using Experienced Consultants

Our CFD Consultants use Ansys Fluent to perform fluid flow simulations. Ansys is an advanced software physics modelling package that incorporates a wide range of applications suitable for fluid dynamics problems (CFD), and fluid-structure interaction problems (FSI), among others. It is well known for its advanced physics modelling and industry-leading accuracy and efficiency.

Even with readily accessible CFD software, whether open-source or commercial, and access to cloud computing solutions, achieving accurate and meaningful results requires industry experience and technical expertise. Incorrect boundary conditions, unsuitable models and settings, or poor mesh quality can lead to misleading results, regardless of the software's accuracy.

Using experienced consultants with technical expertise ensures accurate analysis and reliable simulation results. These consultants possess the skills and experience to select appropriate models, understand complex physics, and generate quality meshes, leading to reliable simulations. Through experimental validation, optimisation, and interdisciplinary insights, they guarantee the accuracy and reliability of the analysis, thereby enabling informed decision-making processes and fostering innovation using computational fluid dynamics.

The use of our CFD consultancy services is cost and time-effective, offering fast turn-around times, efficient and effective analysis of results, and a reduction of the need for large upfront hardware costs and expenses associated with maintaining a full-time team of CFD engineers.

Do you have a project which would benefit from CFD modelling?

Please contact our CFD engineering team to discuss your requirements, and subscribe to our newsletter and follow us on LinkedIn.

 

Next Steps

Site Selector