**CFD** **METHODOLOGY: HOW RENUDA USES CFD FOR YOUR PROJECT**

Starting from a sound understanding of clients’ requirements, Renuda performs end-to-end fluid and thermal flow analyses

to deliver solutions to a wide variety of industrial problems. Results are delivered in the form of:

• optimised designs

• geometry and operating conditions data

• performance parameters

• flow visualisations

• data for further analysis (eg. forces and heat fluxes for mechanical and energy calculations).

The simulation chain is described in the diagram above.

Insightful CFD simulations involve isolating the relevant physical phenomena and carefully defining the problem of interest,

its geometry, and the purpose of the simulations. CFD simulations demand a clear understanding of the physics of a problem

and the numerical modelling required, which includes the capabilities and limitations of physics models and their specific

software implementations.

The simulation process follows this sequence:

**CREATE A VIRTUAL REPRESENTATION**

Unless starting with a clean slate when creating a brand new design, the entry point of the simulation process is usually geometric measurements, data from blue prints or digitised geometric CAD data supplied by the client. The data may need to be joined and cleaned by Renuda before it can be used to create the computational domain. A virtual description of the geometry of interest is then created, based on this data.

**PREPARE SIMULATIONS**

**| ****Define and set up physical and numerical models**

Physical and numerical models are chosen, based on the hypotheses made for the study and the significant physics of the problem. For example, it might not be necessary to solve equations for temperature if temperature changes are thought not to be significant; or turbulence may have a driving influence, so turbulence models must be selected; or the flow Mach number is high and solution algorithms compatible with flow compressibility must be chosen. The choice of models will impact the choice of meshing parameters in the next step.

**|**** Meshing**

Next is the meshing of the surface and the volume of the computational domain. The geometry is discretised into smaller elements, each representing a data point in the solution of the model equations. The higher the number of these elements and associated data, the more the simulations will be able to accurately capture fine spatial variations in the data. This is similar to digital photography, where an image made of different colours, hues and tones is broken into pixels of uniform colour. As the number of pixels increases, the photograph becomes a more accurate representation of reality.

**| ****Impose Operating Conditions**

By imposing boundary conditions on the limits of the computational domains, the operating conditions chosen for the system can be established and the virtual CFD model completed. Boundary conditions may be supplied by the client, or form a set of parametric data in a range that has been specified with the client. A parameter sweep may be set up for studies that aim to establish the performance of a system for different operating points.

**RUN SIMULATIONS**

Calculations are then run, often on multi-processors machines for parallel acceleration, until convergence is obtained for steady-state flows or for a set time for time-resolved, unsteady flows.

**PRODUCE SOLUTIONS THROUGH ANALYSES**

The data from the simulations may now be analysed. Through detailed investigation in 3D of the calculated data and reports on important quantities, which may be local and in real time, or time and space averaged, the main fluid and energy flow features are extracted. Graphs, flow visualisation of streamlines or velocity vector fields, and contour maps of variables such as pressure, temperature and velocity fields are also produced, tracking the evolution of these quantities during the simulations.

In a gradual approach and as part of the process of deriving a solution, the choice of physical models will often be refined along with the meshing specifications during a study, starting with simpler modelling. This can then instruct further detailing, depending on the results of the analysis or as a design becomes final.

>** ****Contact us**** **to discuss your CFD requirements.