What is CFD?
Computational Fluid Dynamics is the simulation of fluid engineering system which is used to analyses the flow over the entire body then to evaluate the performance of the model with the help of initial boundary value problems and numerical methods.
Historically
Applications of CFD
Types of CFD codes
1.Commercial CFD codes
a.FLUENT
b.CFX
c.STAR-CD
d.Gambit
e.CFDRC..etc.
2.Research CFD codes
a.CFDSHIPS-IOWA
3.Public domain softwares
a.PHI3D
b.hydro
c.winpipeD..etc.
4.Grid generation softwares (GRIDGEN GAMBIT)
5.Flow visulation softwares (Tecplot,Fieldview).
Geometric Modeling
Modelling is the mathematical physical problem formulations interms of a continous Intial Boundary Value Problem(IBVP)
IBVP is the form of partial differential equation(PDEs) with appropiate boundary and initial conditions.
Geometric and Domain
Simple geometry: It can be easily created geometries such as circle,plateand pipes.
Complex geometry:It can be created by the PDEs and importing the geometry through the commercial softwares CATIA,Pro-E..etc. using of industrial standards formatssuch as IGES,IGS,STEP,Parasolid,ACIS..etc.
Domain: It is enclosure of the mathmatical model depends on shape and size.
Flow Conditions
Based on the physical phenomena of the mathematical models with different requirements.
For example turbulent flow simulation video
Cooling process
Screw compressor simulations
Wind turbine simulation
Initial Conditions
Based on flow conditions will be either steady or unstaedy flow independently.
1.It is affect the final results then ony affects the convergence path such as no of iterations(steady) and no of time steps (unsteady)
2.For complicated unsteady flow problems.CFD codes are usually run the steady mode for getting the better initial conditions.
Boundary Conditions
It means wheteher ,
slip-free walls or no-slip,periodic,inlet (velocity,constant pressure) and outlet conditions (velocity convective and zero-gradient) and non-reflecting for compressible flows such as acoustics.etc.
Numerical Methods
IBVPs are discretized by the algebraic eguations using numerical methods then assemble the system of algebraic equations and solve the system to get a approximate solutions.
The numerical methods includes
Discretization method
In this method usullay contain three method
a.Finite difference method (only for regular grids)
b.Finite volume method (for irregular meshes)
c.Finite element method (for unstructured meshes or irregular)
Explicit Method (Euler and Runge-kutta method)
It can be easily applied but yield conditionally stable Finite difference equations(FDEs) which are restricted by the time-step
Implicit Method (Beam-warming method)
It is usally unconditionally stable but need efforts on efficiency.Usually higer-order temporal discretization methods are used.
Stability
If it is said to be stable if it does not magnify the errors that apperas in the solution process.
Numerical solvers and parameters
Solvers can be either
a.Direct method-Gauss elimination,cramers rule
b.Iterative method- Gauss seidal and Jacobin methods.
Numerical Parameters
a.convergence limit
b.under relaxation method
c.monitor residuals
d.single or multi precision
e.number of iteration
Grid generation and transformation
Grids can be classified into
a.Structured grids (Hexahedral)
b.Unstructured grids (tetrahedral)
Transformation which means converting the physical domain into the computational domain.
High performance computer postprocessor
CFD computation (eg:3D unsteady flows) are generally very expensive so which requires parallel high performance super computers with the use of multiblock techniques.
Post processing which means visualize the CFD results for example contours,velocity vectors,pathlines,stream lines,strealines and 3D particle flows,iso-surface.
Then CFD UA: To validate and verify by the EFD data.
General Implementation Process of CFD
The follwing step by steps are to execute the entire CFD solution process,
Physics
Mesh
Solvers
Post-Processing
Computational Fluid Dynamics is the simulation of fluid engineering system which is used to analyses the flow over the entire body then to evaluate the performance of the model with the help of initial boundary value problems and numerical methods.
Historically
- Analytical Fluid Dynamics(AFD)
- Experimental Fluid Dynamics(EFD)
- Simulation-based design instead of built&test
- More cost-effective and more rapid than EFD
- Large scale simulation (Eg:Ships and airplanes)
- Environmental effects (wind,weather.etc.)
- Artificial Hazards (radiation,pollution,explosions)
- Physics testing ( planetary boundary layer and stellar evolutions)
Applications of CFD
- Aerospace (aircrafts,missiles,launch vehicles and spaceships and spacecrafts)
- Automotive (lighter passenger and large good carrying vehicles)
- Bio-medical (temperature and natural convections current in the eye flowing laser heating.
- Chemical processing ( reactor vessels and ic engines)
- HVAC (heating ventilation airconditioning system- streamlines for workstation ventilations
- Hydraullics (actuatingthe oil flow into the hydraullic pump)
- Marine (small boats and ships)
- Oil and Gas (flow of oils and gas)
- Power generation (flow arround the cooling towers)
- Sports (helmet,footballs,ruggby).
Types of CFD codes
1.Commercial CFD codes
a.FLUENT
b.CFX
c.STAR-CD
d.Gambit
e.CFDRC..etc.
2.Research CFD codes
a.CFDSHIPS-IOWA
3.Public domain softwares
a.PHI3D
b.hydro
c.winpipeD..etc.
4.Grid generation softwares (GRIDGEN GAMBIT)
5.Flow visulation softwares (Tecplot,Fieldview).
Geometric Modeling
Modelling is the mathematical physical problem formulations interms of a continous Intial Boundary Value Problem(IBVP)
IBVP is the form of partial differential equation(PDEs) with appropiate boundary and initial conditions.
Geometric and Domain
Simple geometry: It can be easily created geometries such as circle,plateand pipes.
Complex geometry:It can be created by the PDEs and importing the geometry through the commercial softwares CATIA,Pro-E..etc. using of industrial standards formatssuch as IGES,IGS,STEP,Parasolid,ACIS..etc.
Domain: It is enclosure of the mathmatical model depends on shape and size.
Flow Conditions
Based on the physical phenomena of the mathematical models with different requirements.
For example turbulent flow simulation video
Cooling process
- Viscous vs invisid (Re)
- External or internal flow (wall bounded or not)
- Compressible vs incompressible (Mach,equation of state0
- Tubulent vs laminar (Re,turbulent models)
- Single or multi phase (Ca,cavitation model and two fluid models)
- Thermal effects/density (Pr,Gr,ec) and conservation of energy
- Free surface flow (Fr) or surface tension flows (We,level-set and surfacr tracking and bubble)
- Chemical reaction and combustion (Pe,Da)
Initial Conditions
Based on flow conditions will be either steady or unstaedy flow independently.
1.It is affect the final results then ony affects the convergence path such as no of iterations(steady) and no of time steps (unsteady)
2.For complicated unsteady flow problems.CFD codes are usually run the steady mode for getting the better initial conditions.
Boundary Conditions
It means wheteher ,
slip-free walls or no-slip,periodic,inlet (velocity,constant pressure) and outlet conditions (velocity convective and zero-gradient) and non-reflecting for compressible flows such as acoustics.etc.
Numerical Methods
IBVPs are discretized by the algebraic eguations using numerical methods then assemble the system of algebraic equations and solve the system to get a approximate solutions.
The numerical methods includes
Discretization method
In this method usullay contain three method
a.Finite difference method (only for regular grids)
b.Finite volume method (for irregular meshes)
c.Finite element method (for unstructured meshes or irregular)
Explicit Method (Euler and Runge-kutta method)
It can be easily applied but yield conditionally stable Finite difference equations(FDEs) which are restricted by the time-step
Implicit Method (Beam-warming method)
It is usally unconditionally stable but need efforts on efficiency.Usually higer-order temporal discretization methods are used.
Stability
If it is said to be stable if it does not magnify the errors that apperas in the solution process.
Numerical solvers and parameters
Solvers can be either
a.Direct method-Gauss elimination,cramers rule
b.Iterative method- Gauss seidal and Jacobin methods.
Numerical Parameters
a.convergence limit
b.under relaxation method
c.monitor residuals
d.single or multi precision
e.number of iteration
Grid generation and transformation
Grids can be classified into
a.Structured grids (Hexahedral)
b.Unstructured grids (tetrahedral)
Transformation which means converting the physical domain into the computational domain.
High performance computer postprocessor
CFD computation (eg:3D unsteady flows) are generally very expensive so which requires parallel high performance super computers with the use of multiblock techniques.
Post processing which means visualize the CFD results for example contours,velocity vectors,pathlines,stream lines,strealines and 3D particle flows,iso-surface.
Then CFD UA: To validate and verify by the EFD data.
General Implementation Process of CFD
The follwing step by steps are to execute the entire CFD solution process,
- Geometry
- Physics
- Mesh
- Solver
- Report
- Postprocessing
- Set the proper co-ordinate system
- Create the appropiate domain which means enclosure shape and size
- If u wants to import from the commercial CAD/CAM softwares,you could import via STD formats.
Physics
- Flow conditions which means folw will be turbulent vs laminar or viscous or inviscous..etc
- Flow properties it will be density,viscosity,thermal conductivity..etc.
- These condition and properties are based on the model dimensional form in industrial norms
- Wheras Non-dimensional means REASERACH CODE ONY
Mesh
- It would be important factor on analyses the mathematical models
- Mesh shoul be well designed to solve the model which depends upon the flow parameter such as Re,We,Ca,Pr,etc..
- Even meshes are generated by the commercial grid generation tools such as GRidgen and gambit.etc.
Solvers
- Set-up appropiate numerical parameters
- Select the relevent solver based on the requirements.
- Solve the problem by using the selection of proper system of equations
- Report saved the time history of the residuals of the velocity,pressure,temperature.etc.
- Reports the integral quantities such as total pressure drop,friction factor (pipe-flow) and lift and drag coefficient (airfoil-flow)
- Plot the graph between the different parametres such as velocity and pressure distribution..etc.
- AFD or EFD can be imported and put it from top of the plots
Post-Processing
- Analyses and visualization of the parameters are vortices,forces,moments,wall shear stress,2D contours and animations.
- To checking the simulaton errors,validations and verification of the simulated numerical parameters.
- To evalute the convergence of the input parameters
















