MER331 – Week 6

An Introduction to Computational
Fluids Dynamics
Professor Anderson
What is CFD?
 the analysis of systems involving fluid flow, heat
transfer, chemical reactions, by means of
computer based simulation
 17th century the “hydraulicists”
 18th and 19th - the “mathematicians”
 1960 - two fold approach – theory and experiments
 Now we have CFD as a tool that complements
theory and experiments (it will never replace
these!)
 CFD is both a research tool and design tool
What is CFD?
 CFD has lagged behind developments in stress
analysis codes
 Powerful computers now make CFD more
accessible
 Advantages of CFD over experiments
 reduction in lead time and costs of new designs
 ability to study systems where controlled
experiments are difficult to perform (large systems)
 ability to study systems under hazardous conditions
at and beyond normal performance limits
 detail of results
Applications
 Aerospace
 Automobile and Engine Applications
 Appliances
 Boats
 Computers
How Does a CFD Code Work?
 CFD codes are structured around the numerical
algorithms that can tackle fluid flow problems

 Three main elements:

1.Pre Processor
2.Solver
3.Post Processor
 1) Pre-Processor
 Consists of the input of
a flow problem to CFD
 User Activities:

 define geometry &

generate grid (50% time)
 selection of phenomena
to be modeled
 definition of fluid
properties
 specification of
boundary and initial
conditions
2) Solver
Three primary numerical solution techniques
 finite difference, finite element, finite control volume

The numerical method performs the following:

 Approximates the unknown variables by simple functions
 Discretization by substitution of the approx-imations into
the governing flow equations and subsequent
mathematical manipulations
 Solution of the algebraic equations
 Solver - Finite Difference Method
 Finite difference methods describe the unknowns f of the
flow problems by means of point samples at the node
points of a grid co-ordinate lines

 Truncated Taylor series expan-

sions are used to generate finite
difference approximations of the
derivatives of f in terms of point
samples of f at each grid point
and its immediate neighbors
Solver - Finite Volume Method
 Based on control volume formulation of analytical fluids
 The domain is divided into a number of control volumes
(aka cells, elements) - the variable of interest is located at
the centroid of the control volume.
 The differential form of the governing equations are
integrated over each control volume.
 Finite difference approximations are substituted for the terms
in the integrated equations (discretization) converts the
integral equations into a system of algebraic equations.
 Set of algebraic equations are solved by an iterative
method.
3) Post Processor
 Provides a “user friendly” (??) way to look at the
results of a simulation
 Domain geometry and grid display
 Vector Plots
 Contour Plots
 Particle Tracking
Problem Solving With CFD
 Results of CFD are at best as good as the physics
embedded in it as at worst as good a its operator
 THESE PROBLEMS ARE COMPLEX
 Prior to running a simulation there is a stage of
identification and formulation of the flow problem in
terms of the physical and chemical phenomena
that need to be considered.
 A successful simulation has
 converged solution
 grid independence
In Lab Today

 We will use StarCCM which is a one of the

major commercial CFD codes.
 Complete the tutorial on Introduction to
STARCCM (handout of online)
 Geometry File is in: C:\Program Files\CD-adapco\STAR-
CCM+7.06.012\DOC\startutorialsdata\introductions\data

 To Pass In: Nothing!