Accueil » Resources » What is CFD simulation?
In this article we will focus on the understanding of CFD simulation in a general way by detailing the different steps common to different types of simulation (HVAC, hydrology, heat transfer, pollution diffusion, fire safety…).
The success of a CFD simulation [computationnal fluids dynamics] necessarily goes through: an understanding of the issues of the model; by the complete description of the geometry of the structure; the development of an adapted network the morphology of the structure, by densifying it in the zones where the gradients of the required quantities are likely to appear; the rigorous study of the boundary and initial conditions, taking into account the most influential aeraulic or hydraulic mechanisms. And finally, a rigorous and critical reading by qualified engineers of the results according to the problem studied.
CFD, an acronym for “Computational Fluid Dynamics”, is an engineering tool that falls under the umbrella of what is known as computer-aided engineering (CAE). More precisely, CFD refers to the simulation of fluid flow, taking into account the physical and chemical phenomena involved (such as turbulence, heat transfer or chemical reactions).
The purpose of wind tunnel testing is to reproduce the interaction between turbulent wind and structures. For structurally stiff structures, it is possible to evaluate the aerodynamic loads on rigid models.
Wind tunnel testing has been widely used for industrial and civil engineering applications over the past five decades .
Wind tunnel testing requires expensive setup and sophisticated instrumentation to measure a range of field variables (wind speed, pressure loads, turbulence intensity, etc.). Its main limitation is that such measures are obtained only at a few specific points of the test section, which significantly restricts overall understanding evolutionary or transient processes of complex unsteady phenomena (such as the vortex shredding, turbulence wakes and thermal stratification).
CFD offers many advantages over wind tunnel testing. In addition to generating real-scale simulations (rather than reduced-scale models for many physical simulations), it also provides complementary data and allows comparison for a given wind of the wind speeds simultaneously between two points. . It is possible to carry out hydrological, aeraulic or thermal studies at different scales: from micro-electronics to building and city studies. The results can be visualized more clearly and explained to as many people as possible.
These methods can be used to solve a very wide range of problems, which we will present below.
Thanks to simulation, the design of a process or a product can be improved without resorting to the construction of prototypes (costly in time and money ); poor decision making can be avoided; a better knowledge of the process or product is obtained, thanks to which it is possible toadvance more quickly in the design process (choice of the best solutions), as well as solve problems that appear in installations or processes already in operation.
Therefore the framework of a physical problem can be posed, it can be studied in digital CFD simulation.
Generally speaking, a fluid simulation project involves a preliminary study of the process / phenomenon to be analyzed, creation of a detailed geometric modelthe selection (and implementation if necessary) of appropriate mathematical modelsthe application of operating data such as boundary conditionsthe numerical calculation (which can vary from a few minutes to a few daysdepending on the complexity of the calculation) and the analysis of the results.
Thus, despite the fact that applications have been developed in recent years to facilitate its use, properly executing a CFD project requires experience and a significant investment in resources.
Before we start a design study, there are a few important questions we need to ask you. These questions are crucial in determining the geometry with which we will begin the analyses, the parts of your design we will focus on, and the parameters we will observe once the analysis is complete.
Once these questions have been answered to improve the understanding of the issues, we detail here the CFD modeling process common to all types of projects.
Effective simulation begins with good modeling techniques both in terms of model integrity and proper creation of different fluid flow regions and mesh optimization. The first step is to design a model for fluid flow analysis. This means modeling the geometry where the flow occurs and optimizing the model for simulation.
To study fluid motion in a design, there must be a model of the flow region. Most 3D models don’t include it by default so it’s a matter of making them from software that complements the original 3D model. On the other hand, it is also a question of preparing the model for the optimization of the mesh in high-stake areas. Thus, we add 3D parts, invisible in the renderings and in the CFD studies which will allow to refine precisely the mesh in the flow areas to be captured in the CFD study.
The generation of the mesh (3D) is an important phase in a CFD analysis, given its influence on the calculated solution. A mesh of very good quality is essential for obtaining a precise, robust and meaningful calculation result.
Before running a CFD simulation, the geometry is divided into small pieces called elements. The corner of each element is a node. The calculation is performed at the nodes. These elements and nodes constitute the mesh.
In three-dimensional models, most elements are tetrahedra: an element with four sides and a triangular face. In two-dimensional models, most of the elements are triangles.
We distinguish between structured and unstructured, orthogonal or free mesh. In a 3D structured mesh , the calculation is done more quickly since it does not require the assembly of a connection matrix. In an unstructured mesh, this is not the case. The advantage of the latter is that it makes it possible to mesh any geometries. On the other hand, the creation and the setting in memory of the matrix can strongly slow down calculation. This type of mesh is used for complex geometries with curves or a large number of elements.
Solid volumes require few elements , unlike fluid volumes which require precise refinement because they cannot move away from a parallelepipedal geometry; indeed for the angles of the very deformed elements, there are risks that the calculation cannot converge.
With regard to the density of the mesh, it is necessary to find a compromise between the cost of the computation time and the desired precision. It is useless to densify the mesh, consequently to increase the number of iterations, if the precision is sufficient with a limited number of elements.
The quality of the mesh has a serious impact on the convergence , the precision of the solution and especially on the computation time. A good mesh quality is based on the minimization of the elements presenting “distortions” and on a good “resolution” in the regions presenting a strong gradient (gap, boundary layers, recirculation, etc.).
The mesh is adapted to be as fine as possible in the critical study areas. This makes it possible to take into account macroscopic phenomena (building volumetry) channeling current tubes by venturi effect whilecorrectly capturing smaller scale aeraulic phenomena (air diffusion).
The initial conditions represent the characteristics of the flow in terms of speed and position of the free surface when starting the simulation. If the calculation starts with random values, the simulation may quickly diverge. In order not to deviate too much from the realistic results and to optimize the calculation time, the initial conditions are studied and chosen before the CFD study.
The study of the boundary conditions is decisive in a modeling, one can summarize the boundary conditions as the hypotheses of the simulation. This is the most critical step for the success of the study, the establishment of the specific boundary conditions of the project should be studied in detail at the beginning of the mission.
The notion of turbulence model is particular in fluid mechanics. It makes it possible to catalog the various structures which coexist in a flow and to give them a certain importance within the flow.
Comparative studies of turbulence models carried out by Combes allowed to designate the model with two transport equations k-ε as the model best suited to generalist flows. It is one of the most used, the most efficient, the simplest and the most widely validated models. k represents the turbulent kinetic energy and ε, the rate of dissipation of the turbulent kinetic energy. Logically we will use it for most of the fluid simulations in thermo-aeraulics and hydrology, but we can select other turbulence models for particular simulations.
The numerical solution is conducted through linearization and discretization of the set of conservation equations, which requires the subdivision of the computational domain into a number of non-contiguous finite volumes (mesh). The resolution of the study consists in the resolution of the non-linear equation system of Navier-Stokes on computer servers dedicated to CFD.
The flow of a fluid in a volume is usually complex and has many low speed recirculations, which makes visualization on a plane difficult. We report the most striking phenomena with maps / sections of situations and very complete explanations.
We have a wide range of representations (current tube, vector fields, isosurface, etc.) which allow us to best transcribe the aeraulic phenomena identified in the technical report.
The interpretation of the results requires the mastery of the CFD analysis software but above all skills in physics and knowledge of the analyzed product in order to explain precisely the different phenomena.
According to our feedback, for the most striking elements, videos are produced showing the different views of the CFD model in a dynamic way. The technical brief that may refer to these videos to make it easier to read. In fact, certain phenomena appearing to be difficult to understand on the map.
These problems are very clearly reduced thanks to the experience acquired on many projects by the engineers of EOLIOS. It is important to carry out this type of study with a qualified team.
Okay, granted, CFD is not the cheapest engineering tool (compared to a standard CAD application or spreadsheet) due to its complexity and requirements (experience, licenses, computational resources).
On the other hand, the results that this type of study offers and their contribution to a design or problem-solving process cannot be compared to those obtained with simpler tools.
The reduction in design time, prototyping savings, and process or product improvement usually outweigh the cost of CFD simulation.
We offer mission protocols adapted to any budget.
If after reading this, you think your project can benefit from CFD tools, do not hesitate to contact us and we will propose you a clear and detailed study protocol.
© EOLIOS 2022 – Tous droits réservés – Mentions légales – Politique de confidentialité | Site réalisé par Erupteo et EOLIOS