Escolar Documentos
Profissional Documentos
Cultura Documentos
AERONAUTICAL
ENGINEERING
Computational
fluid dynamics
REGULATIONS – 2013
VISHNU KUMAR G C
Asst. Prof. – Aero
REG NO:
NAME : _________________________
www.vishnukumar.webs.com nitvishnu@nehrucolleges.com
TEXT BOOKS:
1. Versteeg, H.K., and Malalasekera, W., "An Introduction to Computational Fluid Dynamics: The
finite volume Method", Pearson Education Ltd.Second Edition – 2007.
2. Ghoshdastidar, P.S., "Computer Simulation of flow and heat transfer", Tata McGraw Hill
Publishing Company Ltd., 1998.
68
Page |1 CFD – UNIT I
UNIT I
GOVERNING EQUATIONS AND BOUNDARY CONDITIONS
1.1 Basics of computational fluid dynamics
Computational fluid dynamics (CFD) is a branch of fluid mechanics that uses numerical analysis and data structures to
solve and analyze problems that involve fluid flows. Computers are used to perform the calculations required to
simulate the interaction of liquids and gases with surfaces defined by boundary conditions. It is a branch of continuum
mechanics which deals with numerical simulation of fluid flow and heat transfer problems.
Equating .ds.v =
t .dv
..v.dv = t .dv
..v.dv + t .dv 0
. v 0
t
L.H.S. R.H.S
Total forces = mass x acceleration
(a) Shear forces: = density × volume × acceleration
dxdydz
D
In XY plane zx zx dz zx dxdy (U )
z Dt
yx
In XZ plane yx dz yx dxdz
z
In YZ plane xx xx dz xx dydz
z
P
Pressure P P dx dydz
z
(b) Body force = mass x force component
= density x volume x force
dxdydz f x
xx yx P D
dxdydz dxdydz zx dxdydz dxdydz dxdydz f x dxdydz (U )
x y z z Dt
xx yx zx P DU
fx
x y z z Dt
dxdydz f x v
u zx
In XY plane u zx dz u zx dxdy
z
L.H.S
u yx
In XZ plane u yx dz u yx dxdz
Rate of change of energy = internal energy + Kinetic energy z
v2 u xx
dxdydz e In YZ plane u xx dz u xx dydz
t 2 z
uP
Pressure uP uP dx dydz
z
Equating
o o o
v2 o
x q q q z u xx u yx u zx
e 2 = P q x y z + f x v +
y
t x y z
1.3 Chemical species transport
Rate of change of mass of + Convection + diffusion = Rate of generation of chemical species per unit volume
ml
div( uml J ) R1 R – Generated by chemical reaction (if no reaction =0)
t
In order to obtain unique solution of a PDE, a set of supplementary conditions must be provided to determine the
arbitrary functions which result from the integration of PDE.
(a) Initial condition: It is a requirement for which the dependent variable is specified at initial state.
(b) Boundary condition: It is a requirement that the dependent variable or its derivative must satisfy on the
boundary of domain.
Turbulent Flow. In turbulent flow, the velocity components and other variables (e.g. pressure, density - if the fluid is
compressible, temperature - if the temperature is not uniform) at a point fluctuate with time in an apparently
random fashion.
In general, turbulent flow is time-dependent, rotational, and three dimensional. For instance, measurement of the
velocity component v1 at some stationary point in the flow may produce a plot. The velocity can be regarded as
consisting of an average value V1 indicated by the dashed line, plus a random fluctuation v1’.
The over score denotes average values and the prime denotes fluctuations. Turbulence will always occur for
sufficiently high Re numbers, regardless of the geometry of flow under consideration. The origin of turbulence rests
in small perturbations imposed on the flow; for instance, by wall roughness, by small variations in fluid density, by
mechanical vibrations, etc. At low Re numbers such disturbances are damped out by the fluid viscosity and the flow
remains laminar, but at high Re (when convective momentum transport dominates over viscous forces) they can grow
and propagate, giving rise to the chaotic phenomena perceived as turbulence. Turbulent flows can be very difficult to
analyze.
The first equation represents kinetic energy of the mean flow (MKE) and second equation represents kinetic energy
of turbulent part of the flow (TKE). The second equation is also-called turbulent kinetic energy. Turbulent kinetic
energy is one of the most important variables in micrometeorology. It’s a measure of the intensity of turbulence. But
since the turbulence may change in time we are more interested in TKE budget equation (∂e/∂t).
The governing equations of fluid flow are second order partial differential equations. Choice of the numerical
solution technique and number of initial/boundary conditions required for their solution depends on their
mathematical character. Since these equations are system of coupled with independent variables, their
mathematical classification is difficult.
Mathematical classification of quasi linear PDE is usually based on the classification procedure which is evolved for
second order PDE in two independent variables(x, y).
Where A, B, C, D, E, F and G are functions of x and y. Classification is based on the roots of the associated characteristic
equation.
UNIT II
FINITE DIFFERENCE AND FINITE VOLUME METHODS FOR DIFFUSION
2.1 DERIVATIONS OF FINITE DIFFERENCE EQUATIONS:
f (x) 2 2 f (x)3 3 f
f ( x x) f ( x) (x) ...
x 2! x 3! x3
Central Difference
f (x) 2 2 f f (x) 2 2 f
f ( x x) f ( x) (x) f ( x x) f ( x) (x)
x 2! x x 2! x
f (x) f
2 2
f ( x x) f ( x) x)
x 2! x
f (x)2 2 f f (x)2 2 f
f ( x x) f ( x x) f ( x) (x) f ( x) (x)
x 2! x x 2! x
f
f ( x x) f ( x x) 2(x)
x
f ( x x) f ( x x) f
2(x) x
f (x) 2 2 f f (2x) 2 2 f
f ( x x) f ( x) (x) f ( x 2x) f ( x) (2x)
x 2! x x 2! x
f (x) f
2 2
f (x)2 2 f f ( x 2x) f ( x) 2x
2 f ( x x) 2 f ( x) 2(x) 2 x 2 x
x 2! x
Forward Difference
f 2 f f 2 f
2 f ( x x) f ( x 2x) 2 f ( x) 2x (x) 2 f ( x) 2x 2(x) 2
x x x x
2 f
2 f ( x x) f ( x 2x) f ( x) (x) 2
x
2 f ( x x) f ( x 2x) f ( x) 2 f
(x)2 x
2 f ( x x) f ( x 2x) f ( x) 2 f 2 fi 1 fi 2 fi 2 f
(x)2 x (x) 2 x
Backward Difference
f 2 f f 2 f
2 f ( x x) f ( x 2x) 2 f ( x) 2x (x) 2 f ( x) 2x 2(x) 2
x x x x
f 2 f f 2 f
2 f ( x x) f ( x 2x) 2 f ( x) 2x (x) 2 f ( x) 2x 2(x) 2
x x x x
2 f
2 f ( x x) f ( x 2x) f ( x) (x) 2
x
2 f ( x x) f ( x 2x) f ( x) 2 f 2 fi 1 fi 2 fi 2 f
(x)2 x (x)2 x
Central Difference
f (x)2 2 f f (x) 2 2 f
f ( x x) f ( x) (x) f ( x x) f ( x) (x)
x 2! x x 2! x
f (x) 2 2 f f (x) 2 2 f
f ( x x) f ( x x) f ( x) (x) f ( x) (x)
x 2! x x 2! x
2 f
f ( x x) f ( x x) 2 f ( x) (x) 2
x
f ( x x) f ( x x) 2 f ( x) 2 f fi 1 2 fi fi 1 2 f
(x) 2 x (x) 2 x
Step 2: Discretization
Integration of governing equation over a control volume to yield a discretised equation at its nodal point P.
Discretised equations will be setup at each of the nodal points Boundary conditions will be applied to
transform it into linear algebraic equations which will be solved in matrix form.
Example
Consider the problem of source free heat conduction in an insulated rod whose ends are maintained at constant
temperatures of 100C and 500C respectively. Calculate the steady state temperature distribution in the rod.
Thermal conductivity k equals 1000W/m/K. cross sectional area is 10x10-3m2.
Step 1 : Dividing the length of the rod into five equal control volumes which gives ∆x=0.1m
Discretized equation
For Nodes 2,3,4:
For Node 1:
For Node 5:
Rearranging this
Formulation with one unknown in terms of the Formulation with more than one unknown in each FDE.
known values. Resulting equation at time level n+1 Stability will be achieved with larger time step is permitted.
is solved independently to provide the values of the
unknowns. The boundary condition lag behind the
computataion by one step.
Forward time central space method Crank – Nicolson method
Finite-element method: It can locally refine the mesh close to that corner; the element density
increases locally. The finer the mesh (i.e., more elements), the more accurate approximation one gets
for the stress field around the corner of interest. All kinds of structural analysis, heat transfer, chemical
engineering, electromagnetics, multi-physics, and CFD.
Finite-difference method: Boundary conditions are needed to truncate the computational domain.
Weather calculations, astrophysics, seismology, physical realism in computer graphics, and special
effects.
Finite-volume method: The finite-volume method’s strength is that it only needs to do flux evaluation
for the cell boundaries. CFD, heat transfer, and chemical engineering.
UNIT III
FINITE VOLUME METHOD FOR CONVECTION AND DIFFUSION
3.1 STEADY ONE DIMENSIONAL CONVECTION AND DIFFUSION:
In the absence of sources, the steady convection and diffusion of a property Ф in a one dimensional field is given by
For general node P the neighboring nodes as W and E and the control volume faces by w and e.
F – Convective mass flux per unit area, D- diffusion conductance at cell faces.
Final Equation
where
3.2.1 CONSERVATIVENESS
Consider four control volumes and applying central differencing to calculate the diffusive flux across the cell faces.
Flux leaving the west face: Flux entering the east face:
Overall flux over control volume will be: Flux cancels out, since . Flux
consistency ensures conservation property over the entire domain.
3.2.2 BOUNDEDNESS
Diagonal dominance is a desirable feature for boundedness which states that in the absence of sources the internal
node of the property should be bounded by its boundary values.
3.2.3 TRANSPORTIVENESS
The transportive property of fluid flow (Roache, 1996) can be defined by cell Peclet number as a measure of relative
strengths of convection and diffusion.
Rearranging in terms of coefficients of P,W,E terms Rearranging in terms of coefficients of P,W,E terms
In a pure convection process, without physical diffusion the error is largest for the coarse grid and refining the
grid can overcome the problem of false diffusion.
Central coefficient
For steady one dimensional convection-diffusion term as
UNIT IV
FLOW FIELD ANALYSIS
4.1 FINITE VOLUME METHODS:
Defining velocity and pressures at the nodes of the control volume, a highly non uniform pressure field can act like a
uniform field. The pressure at the node P is not defined in equations which leads to spatial oscillations. Evaluating
scalar variables such as pressure, density, temperature at ordinary nodal points and velocity components on
staggered grids centered around cell faces.
The discretised u-momentum equation for the velocity at location (i, J) is given by
Where
At each iteration level the values of F are computed using the u and v velocity components resulting from the
previous section.
The pressure and velocity components needs to be updated at every iteration to achieve convergence.
U momentum equation
v momentum equation
UNIT V
TURBULENCE MODELS AND MESH GENERATION
5.1 TURBULENCE MODELS
A turbulence modelling is a computational procedure to close the system of mean flow equations
5.3 K- ξ model
The thin shear layers changes in the flow direction and the influence of turbulence on the mean flow can be
analysed. It focus on the mechanisms that affect turbulent kinetic energy.
GRID
The computational domain is selected to be rectangular in shape where the interior grid points are distributed
along the grid lines.
5.4 STRUCTURED GRID
If the grids are aligned with the grid lines with good mapping and uniform spacing it is structured grid.
Physical domain Computational domain
Advantages
1. Computationally they are fast
2. Metrics may be evaluated analytically, avoiding numerical errors
3. Ability to cluster grid points in different regions which can be easily implemented
Disadvantages:
1. Discontinuities at a boundary may propogate into the interior region which could lead to errors due to
sudden change in the metrics.
2. Control of grid smoothness and skewness is difficult task.
1. A mapping which makes one to one correspondence ensuring grid lines of the same family which do not cross
each other.
2. Smoothness of the grid point distribution.
3. Orthogonally or near orthogonally of the grid lines.
4. Options for grid point clustering.
DELAUNAY METHOD
MESH REFINEMENT
Consider the grid of a cross section. The outer boundary of the free stream will be far away to capture the flow
properties around it.