6/6/2012

Fluid Dynamics and Heat Transfer

E-Outreach Workshop • Ultimate Objective: Determination of velocity
on field, pressure field, temperature field, density
Computational Fluid Dynamics field.
– Determination of quantities of engineering interest:
The National Mission on Education Through ICT different types of forces (lift/drag), heat transfer
(MHRD, Government of India) coefficient, etc.
– Determination of other derived flow fields, e.g.,
June 12 – 22, 2012 vorticity field.
Indian Institute of Technology Bombay
• Analytical (theoretical) or experimental or
Course Coordinators Project Coordinator Project Manager computational (numerical) approaches to
Bhalchandra Puranik Professor D. B. Phatak Mukta Atrey achieve the objective.
Atul Sharma

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

Investigation Approaches Investigation Approaches

• Analytical (theoretical) • Experimental
 Governing equations/mathematical models  Dimensional analysis/model studies.
(Conservation of mass, momentum and energy).
 Measurement of relevant quantities (velocity,
 Sometimes additional equations are needed: equation
of state, turbulence closure, chemical reactions, etc. pressure, temperature, etc.).
 Analytical approach => “Closed-form” solutions.  Analysis of measurement data – flow field
 Often times requires use of advanced mathematical information.
techniques.  Capable of being most realistic.
 Limited to simple geometrical and physical situations  Equipment issues, scaling issues,
– restricted use.
measurement issues.
 Time consuming, and can be very expensive.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

1
 6/6/2012

Investigation Approaches The “CFD(HT) Activity”

• Computational (numerical) One description:
 Use of a computer to solve the governing equations. “Activity where the integral/differential governing
 “Number crunching”, i.e., solution obtained in terms of equations of fluid flow (and heat transfer) are
numbers. replaced with approximate algebraic equations
 Analysis of solution (plotting, etc.). that are solved within the domain of interest to
 Can handle complicated geometries and physics. obtain the relevant flow field information on a
 Truncation errors, model limitations, issues with discrete basis”.
numerical schemes, computational cost (still an issue
=> Final outcome of a CFD analysis is a collection
in some cases).
of numbers.
 Very affordable and hence highly popular, in recent
times!!

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

Why CFD? Objectives of the Course

• Research • Introduction of Fundamentals of Computational Fluid
 Complementary tool to experiments and theory. Dynamics and Heat Transfer (CFDHT).
 Valuable flow field information that experimental – Level of material: advanced undergraduate/beginning
measurements may not be able to provide. postgraduate.

• Design and Analysis of systems • Review of essential fluid dynamics: background for
 Cost-effective tool to run “what-if” scenarios to arrive at CFDHT.
optimum designs. • Introduction to the methods/algorithms that are working
• Companion tool in teaching Fluid Dynamics “behind the screen” when a CFD software operates.
and Heat Transfer – Focus on “numerical methods” and not on “numerical analysis”.
 Appreciation of reality by students. – Model problems and their numerical solutions => Development of
codes (open source programming environment “Scilab”).
 Generation of interest in Fluid Dynamics and Heat
Transfer phenomena. • No CFD software will be introduced.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

2
 6/6/2012

Expected Outcome An Example1

• Understanding the governing equations: origin,
derivations, manipulations.
• Understanding the basic techniques in CFDHT.
• Development of capability to write computer Free Stream
Mach number = 9.6
codes to appreciate the working of a CFD
software.
• Generating a knowledge base to formulate and
teach CFDHT course(s) in your program(s). High Mach number flow past a “Double Cone”
(Viscous, compressible flow with complex features)

1Praveen Nair, A Solution-Mapped High-Resolution Scheme for Hypersonic Heat Flux Prediction,
Ph. D. thesis, IIT Bombay, 2011.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

An Example1 An Example1

Supersonic
Exit
ld
r Fie
Fa

al l
W
Nitrogen
P∞=36.05Pa,
T∞ = 185.6 K
M∞=9.59

“Structured” Mesh/Grid “Unstructured” Mesh/Grid

Problem Setting

1Praveen Nair, A Solution-Mapped High-Resolution Scheme for Hypersonic Heat Flux Prediction, 1Praveen Nair, A Solution-Mapped High-Resolution Scheme for Hypersonic Heat Flux Prediction,
Ph. D. thesis, IIT Bombay, 2011. Ph. D. thesis, IIT Bombay, 2011.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

3
 6/6/2012

An Example1

Mach number contours

PART I
Essentials of Fluid Dynamics
Bhalchandra Puranik

1Praveen Nair, A Solution-Mapped High-Resolution Scheme for Hypersonic Heat Flux Prediction,
Ph. D. thesis, IIT Bombay, 2011.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

Topics (Part I) Reference Books (Part I)

• Introduction and Mathematical background. – M. C. Potter, D. C. Wiggert, “Fluid Mechanics”, India Edition,
Cengage Learning, 2010.
• Fluid Dynamics: Integral analysis. – V. Gupta, S. K. Gupta, “Fluid Mechanics and Its Applications”, 2nd
• Kinematics of fluid motion. Edition, New Age International, 2011.
– R. W. Fox, P. J. Pritchard, A. T. McDonald, “Introduction to Fluid
• Fluid Dynamics: Differential analysis. Mechanics”, 7th Edition (SI Version), Wiley India Edition, 2010.
• Some analytical solutions of the governing – M. L. Boas, “Mathematical Methods in the Physical Sciences”, 2nd
Edition, Wiley, 2002. (Indian Edition)
equations.
– M. C. Potter, J. L. Goldberg, E. F. Aboufadel, “Advanced
• Introduction to numerical solution methodology Engineering Mathematics”, Third Edition, Oxford University
Press, 2005. (Indian Edition)
using the Finite Difference Method.
– J. D. Anderson, “Computational Fluid Dynamics: The Basics with
Applications”, International Edition, McGraw-Hill, 1995.
– K. A. Hoffmann, S. T. Chiang, “Computational Fluid Dynamics for
Engineers – Volume I”, Engineering Education System, 1993.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

4
 6/6/2012

Introduction - 1 Introduction - 2
• Scope of Fluid Mechanics: Equilibrium and • Fluids: Liquids and Gases.
motion of fluids. • Microscopic (Molecular) or Macroscopic
approaches:
• Applications: Various branches of Science – Molecular: “Follow each molecule”; Practically,
statistical methods.
and Engineering; Natural phenomena; – Macroscopic: “Overall/average behavior”; Continuum
Sports. model.
• Continuum model:
• Fluid*: When at rest, cannot resist a shear – Continuous distribution of fluid matter and the
force, no matter how small => no static associated properties.
– Disregard the discrete molecular structure.
equilibrium under the action of shear force. – Use methods of differential calculus (point-wise smooth
distribution of fluid properties).

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

Introduction - 3 Introduction - 4
• Continuum model: • Important physical properties:
– Validity of continuum model? – Pressure (p): molecular origin, compressive in
– Concept of a “fluid point” => statistically consistent nature.
values of properties (e.g., density) in an
infinitesimally small volume. – Density (r): mass per unit volume.
– The mean free path (l) << Linear dimension of the – Specific weight ( = rg): weight per unit volume.
“fluid point volume” << Characteristic dimension of
the problem (L). – Specific volume (v=1/ r):volume per unit mass.
– The Knudsen number, Kn = l /L. – Dynamic viscosity (m):characterization of the
– If Kn < 0.001 (approx.), continuum model is valid. resistance offered by a fluid in motion to the
– Invalid for (i) highly rarefied flows, (ii) applied shear forces.
micro/nanoscale flows.

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

5
 6/6/2012

Introduction - 5 Introduction - 6
• Dynamic viscosity: • Dynamic viscosity:
– A simple shear flow: only u=u(y) – t(stress set-up in the fluid to oppose the
du applied shear stress).
Q’
y
Q u y
dy – In general n
y  du 
x P u    
P’
 dy 
– Rate of shear strain (segment PQ): – If n=1, we have a Newtonian fluid, and
 du  
  u  y t  ut   du 
 l  l
1 QQ ' PP '  1  dy      
t  0      du  dy 
t  l PQ  t  y  dy
 
 
E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

Introduction - 7 Introduction - 8
– Taylor series expansion (function of one variable) – Vector products and differentiation rules
2 3  
df d 2 f x  d 3 f x  A  B  Ax Bx  Ay By  Az Bz  AB cos 
f ( x  x)  f ( x)  x   2  3     
dx dx 2! dx 3! A  B  AB sin 
– Taylor series expansion (function of two variables)
iˆ ˆj kˆ
f f  
f ( x  x, y  y )  f ( x, y )  x   y   A  B  Ax Ay Az  iˆ Ay Bz  Az B y   ˆj  Az Bx  Ax Bz   kˆAx B y  Ay Bx 
x y
Bx By Bz
  2 f x 2  2 f y 2  2 f xy   
 2   2    d  da  dA
 x 2!

2
y 2! xy 2!  dt
 
aA 
dt
Aa
dt
 
– Leibnitz rule (differentiation of an integral) d   dA   dB
v( x ) v( x ) dt

AB  dt
 B  A
dt
d f dv du  
f ( x, t ) dt   dt  f ( x, v)  f ( x, u )   dA   dB
dx u ( x )
d
u( x) x dx dx
dt

A B  dt
 B  A
dt

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

6
 6/6/2012

Introduction - 9 Introduction - 10
– Fields – Gradient, divergence.
• “Field” => continuous distribution of a physical quantity •  is normal to the surface   constant.
represented using appropriate functions of space (and time).
•  represents the largest spatial rate of change of  .
• Scalar field: Temperature field, T  T  x, y , z  .
  • Vector operator  :
• Vector field: Velocity field, V  V  x, y, z , t.
– Directional derivative and gradient  ˆ  ˆ 
  iˆ  j k
• Rate of change of a field variable with respect to distance in a x y z
given direction => directional derivative. • If  is a scalar,  is a vector.
• Let     x, y , z  be a field. The directional derivative in a
direction s • Divergence of a vector field (a scalar).
d  A A A
   sˆ  A  x  y  z
ds x y z
 ˆ  ˆ   2  2  2
  iˆ j k Gradient of  •      2  2  2    2
x y z x y z

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay

Introduction - 11 Introduction - 12
– Gradient, divergence, curl. – Integral theorems of use in the present course material
• Curl of a vector field (a vector). A n̂ Unit vector, normal to the
iˆ ˆj kˆ
  surface A, +ve if pointing
 
 B  outward from the enclosed
x y z
Bx By Bz
 volume.

•      0

•  
   V  0
   Gauss’ theorem Divergence theorem
•   
 

    V     V   2V

 
•  
  V    V    V  d   nˆdA    Bd   B  nˆdA
    A  A
•  
  V    V    V
 scalar

E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay E-Outreach Workshop on CFD Bhalchandra Puranik, MED, IIT Bombay